当前位置：文档库 › 可靠性增长

可靠性增长

Reliability assessment of photovoltaic power systems:Review of current status and future

perspectives

Peng Zhang a ,?,Wenyuan Li b ,Sherwin Li a ,Yang Wang a ,c ,Weidong Xiao d

Department of Electrical and Computer Engineering,University of Connecticut,Storrs,CT 06269-2157,USA b

BC Hydro and Power Authority,Vancouver,BC,Canada V7X 1V5c

Chongqing University,Chongqing 400030,China d

Electrical Power Engineering Program,Masdar Institute of Science and Technology,Abu Dhabi,United Arab Emirates

a r t i c l e i n f o Article history:

Received 1October 2012

Received in revised form 1December 2012Accepted 3December 2012

Available online 2January 2013Keywords:

Photovoltaic power system Reliability assessment

Photovoltaic energy integration Active distribution network Cybersecurity

Voltage control scheme

a b s t r a c t

Quantitative reliability assessment of photovoltaic (PV)power system is an indispensable technology to assure reliable and utility-friendly integration of PV generation.This paper reviews the state-of-the-art technologies for evaluating the reliability of large-scale PV systems and the effect of PV interconnection on the reliability of local distribution system.The discussions are extended to emerging research topics including time varying and ambient-condition-dependent failure rates of critical PV system components,accurate operating models of PV generators in both interconnected and islanded modes,and the reliabil-ity evaluation of active distribution networks with PV penetration and transmission level Giga-PV sys-tem.A vision for the future research is presented,with a focus on the cyber-physical perspective of the PV reliability,modeling of PV voltage control scheme for reliability assessment,reliability assessment for PV systems under extreme events and PV reliability assessment considering cybersecurity.

Contents 1.Introduction:why is PV reliability assessment important?....................................................................8232.Current status and challenges...........................................................................................8233.

Basic aspects in reliability evaluation of grid-connected PV systems............................................................8243.1.Reliability evaluation of critical components in PV system ..............................................................

8253.1.1.PV modules.............................................................................................8253.1.2.Inverter ................................................................................................

8263.2.Reliability evaluation methods of PV system .........................................................................8273.3.Reliability indices for PV system ...................................................................................8284.

Future perspectives on PV reliability assessment ...........................................................................8284.1.Cyber-physical system perspective on reliability assessment of PV system .................................................8284.2.Modeling voltage control scheme in reliability assessment..............................................................8294.3.PV reliability assessment under extreme weather conditions ............................................................8294.4.PV reliability assessment considering cyber vulnerability,attack and security...............................................8305.

Reliability evaluation of power grids with PV systems .......................................................................8305.1.Active distribution network including PV microgrids...................................................................8305.2.Reliability for future Giga-PV system connected to power transmission grid................................................8316.

Conclusions..........................................................................................................831References ..........................................................................................................

831

?Corresponding author.Tel.:+18604863075;fax:+18604862447.

E-mail addresses:peng@https://www.wendangku.net/doc/dc5930730.html, (P.Zhang),wen.yuan.li@https://www.wendangku.net/doc/dc5930730.html, (W.Li),sherwin.li@https://www.wendangku.net/doc/dc5930730.html, (S.Li),wangyanghh@https://www.wendangku.net/doc/dc5930730.html, (Y.Wang),mwxiao@mansdar.ac.ae (W.Xiao).

1.Introduction:why is PV reliability assessment important?

Electricity generated from photovoltaic(PV)power systems is a major renewable energy source which involves zero greenhouse gas emission and no fossil fuel consumption.The total capacity of grid-connected PV power systems has been grown exponentially from300MW in2000to about67GW in2011[1].This capacity, however,is not?rm because of the unreliable nature and probabi-listic behavior of PV power systems.

Relatively high risks exist both inside and outside of PV power systems[2].High uncertainty and variability associated with the system components and environmental factors pose major chal-lenges in designing large PV power system[3].First,a PV power system is composed of many vulnerable components[4,5]whose lifecycle reliability is highly susceptible to temperature,power losses,and ambient environments.Meanwhile,solar insolation and power input of PV system are highly variable and uncontrolla-ble;leading to high electrical stress in PV panels that may shorten the operational lifecycles and power electronic interfaces and con-sequently results in lower system reliability compared to conven-tional generation sources.Second,high penetration of PV generation will bring detrimental effect to power distribution net-work.Signi?cant reverse power?ow may cause unacceptable volt-age rise on distribution feeder.Overvoltage may trigger the protection in PV inverters,which as a result will shut down PV gen-eration,causing sudden change in power?ow and abrupt voltage ?uctuation.Reverse power?ow and voltage?uctuation may also increase the number of operations of on-load tap changers(OLTCs), which will shorten the useful lives of transformers.Distribution networks connected with PVs,therefore,have a high risk for in-creased maintenance costs and power outages,which necessitate methodologies and tools to quantify the reliability of grid-con-nected PV systems.

Risk assessment is of fundamental importance for planning and operation of both PV power systems and PV-connected distribution networks.Its major utilities include:

(1)Quantifying risks in PV systems and choosing optimal PV

system design.

(2)Determining effective measures to mitigate risks.

(3)Justifying acceptable PV penetration level in distribution

network.

(4)Probabilistically evaluating the impacts of intermittent PV

resources on power system adequacy,security,spinning reserve,planning and real-time operation.

(5)Designing recon?gurable distributed energy storage to

leverage PV application.

(6)Finding planning and operational solutions to address the

challenges of high penetration of PV to distribution network in a least cost manner while achieving the maximum level of reliability.

Risk assessment of PV power systems,therefore,is an indis-pensable technology that assures reliable PV generation integra-tion.Practical applications of PV risk assessment theory will bring direct and indirect bene?ts for both utility companies and customers including increased revenue,higher energy yield,im-proved power quality,extended equipment operational life and less carbon emission.

2.Current status and challenges

Increasing attention is being paid to PV system reliability in re-cent years due to rapid growth of PV power installation in residen-tial[6]and commercial buildings as well as military bases.Cost-reduction in production of PV modules together with eco-nomic incentives offered by government will further increase the installed capacity of solar power in the foreseeable future.Failures in PV systems,therefore,will result in signi?cant amount of eco-nomic losses[7].The reliability of grid-connected PV power sys-tems has been of great concern to both utility companies and customers[8].

Although the PV reliability issue was already identi?ed three decades ago[9],reliability quanti?cation of an entire PV genera-tion station remains unresolved due to the complex nature of PV systems.The existing literature mostly focuses on reliability assessment for the power electronic components such as IGBT [10],capacitor[11]and inverter[12,13],whereas much fewer ref-erences discuss the reliability evaluation for entire PV system.Refs. [14,15]presented simpli?ed,system-level models for PV system reliability using a Markov modeling concept.Hierarchical Reliabil-ity Block Diagram was developed[16]to model the behavior of PV system.Ref.[17]quanti?ed the impact of inverter failures on total lifetime of PV system using Monte Carlo simulation.Ref.[18]pro-posed Latin Hypercube Sampling(LHS)technique to integrate strati?ed and random sampling in order to improve its computa-tional speed for obtaining the reliability indices.In the above liter-ature,failure rates of electronic elements in a PV system are treated as constants.These parameters,however,actually vary with system states including solar insolation[19],ambient tem-perature[20],and load level[21].The unrealistic assumptions in reliability analysis may give inaccurate or misleading results.For instance,it was once concluded that‘‘capacitors’contribution to failure rate is quite small’’[22],which may be inconsistent with observations from industrial practice.

On the topic of grid integration of PVs,the National Renewable Energy Laboratory(NREL)has conducted extensive surveys to ex-plore the impact of high penetration PVs on power system plan-ning and operation[23].It has been identi?ed that PV integration is closely tied to overall distribution system reliability [24].Recently,a framework,which is based on Markov reward models(MRM),is proposed to integrate reliability and perfor-mance analysis of grid-tied PV systems[25].This proposed frame-work may help understand the trade-off between repair policies and replacement/overhaul costs.In addition,the effect of reactive power shortage on the distribution network with high PV penetra-tion has been studied[26].Hence,it is obvious that reliability assessment theory suitable for distribution systems integrated with PV generation has become a highly needed technology to build a high-penetration renewable energy future.In the era of smart grid,the microgrid is a mainstream solution for grid integra-tion of PV systems.Reliability evaluation of active distribution sys-tems including PV microgrids becomes a major technical challenge to be tackled.

In previous work,the microgrid was often treated as a small sized conventional power grid where the failure modes of power electronic interfaces were not considered in microgrid reliability evaluation[27–31].These methods may be practical for estimating microgrids with combined heat and power plants(CHPs)[32]or conventional generators,but are not suitable to analyze a distribu-tion network with PVs or other renewable sources.The effect of converter topologies is incorporated into reliability evaluation of DC microgrid by the use of minimum cut sets[33].This approach, however,neither considered the impact of power losses and ambi-ent condition on converter reliability nor can be extended to distri-bution system reliability assessment.Ref.[34]has pointed out that modeling the operation mode transitions is a major challenge in reliability evaluation of microgrids.Reliability of PV/wind micro-grid operating in an islanded mode was studied using Monte Carlo simulation[35].Again,this approach only dealt with input power of PV array without considering the reliability of PV inverters.Fault

P.Zhang et al./Applied Energy104(2013)822–833823

Tree Analysis (FTA)has been used to evaluate the reliability of islanded microgrid in emergency mode [36].The limitation is that this FTA approach can only compute small-scale systems and can-not deal with interconnected modes.It has been realized that a multi-state model is needed for modeling PV generators due to the intermittent nature of solar radiation [37].This method,how-ever,neither considered the impact of input power and tempera-ture on system reliability nor modeled islanded modes of PV microgrids.An analytical approach was proposed [38]to study the effect of distributed generators (DGs)on distribution reliability,where the DG outputs,DG failures and load variations were con-sidered.An event-based Monte Carlo method was developed [39]to evaluate the effect of intentional islanding and switching oper-ations on distribution reliability.Furthermore,pseudo-sequential Monte Carlos has been adopted for the reliability of the active dis-tribution networks [40].The former approach is unable to deal with ?exible operation modes of microgrids,and the latter as-sumes constant loads and DG outputs under islanding situations without considering the intermittent features of DGs.

In summary,the following technical issues either remain unre-solved or are still under further investigation:

(1)Developing power-input/power-loss/temperature-depen-dent failure rates for power electronic components in PV

systems;

(2)Incorporating a power input curve and PV voltage regulation

schemes into PV reliability assessment;

(3)De?ning PV reliability metrics to quantify energy availability

and outage time;

(4)Building the multi-state model of PV microgrid by using reli-ability results of PV systems;

(5)Developing new reliability evaluation algorithms to evaluate

active distribution systems with embedded PV microgrids.3.Basic aspects in reliability evaluation of grid-connected PV systems

This section offers a systematic and detailed summary of PV reliability evaluation technologies recently developed,including practical approaches to quanti?cation of the effects of input power and ambient conditions on failure rates of critical PV components,basic reliability evaluation methods of PV systems and reliability indices for PV reliability indicators.

Large-scale grid-connected PV systems are usually connected either in a centralized or a string/multi-string structure,as shown in Figs.1and 2respectively.The distinguishing feature of the string inverter system is that each string has its own inverter to convert DC electricity to AC output.If a centralized system has the same total capacity as an n -string-inverter system,the capacity of each string inverter is only one-n th of that of the central inver-ter.Another PV topology is the micro-inverter system [41–43].In this structure,the micro-inverter and the PV panel are integrated as one electrical device,which is directly connected to distribution grid through an AC bus,as shown in Fig.3.The purpose of the mi-cro-inverter system is to achieve high modularity,easy installation and enhanced safety.In addition,the maximum power point (MPP)of each module can be tracked individually by the corresponding inverter.Hence,this topology has a potential to better

optimize

824P.Zhang et al./Applied Energy 104(2013)822–833

the PV power generation under partial shading conditions,com-pared to the other topologies.At the same time,micro-inverter system may also improve reliability by reducing converter temper-ature and eliminating electrolytic capacitors.

3.1.Reliability evaluation of critical components in PV system 3.1.1.PV modules

PV modules (also PV panels)are the packaged,connected assembly of PV cells,which are often considered as the most reli-able elements in PV systems.A typical PV module converts 4–17%of the solar insolation into electricity [44–47].The conver-sion ef?ciency mainly depends on solar insolation,operating tem-perature and electrical load [48,49].Nevertheless,the modules can also fail or degrade in their long-term lifecycle [50,51].In the past years,therefore,there were many researches primarily focusing on the reliability of PV modules.In [17],various degradation and fail-ure modes of PV modules are presented.The paper also develops a procedure to assess the degradation,failure modes,as well as their effect on PV module parameters.Ref.[52]proposed to characterize the degradation effect in terms of maximum power point (MPP)and lost hours due to dust accumulation.However,

P.Zhang et al./Applied Energy 104(2013)822–833825

experimental results are needed to validate this https://www.wendangku.net/doc/dc5930730.html,ing proba-bility methods,Ref.[53]proposes a mathematical degradation model for reliability predication of PV module.The model is based on the assumption of linear degradation of reliability parameters and Gaussian distribution of PV power outputs.

Topology is another important aspect associated with PV mod-ule reliability.The researches on the topology of PV modules can be traced back to1980s and even earlier[54,55].Recently,the net-work reliability theories are used to explore reliability of large-scale PV modules.For example,the minimum cut-set technique is used in[56]to investigate the reliability of several different con-nection modes of PV modules,i.e.,the series,parallel,series–paral-lel,total-cross-tied,bridge-linked,and their different combinations.By applying cut-set technique,it was found that to-tal cross-tied(TCT)and bridge-linked(BL)con?gurations increase the operational lifetime of the PV arrays by30%.

3.1.2.Inverter

As made by semiconductor modules,inverters are among the vulnerable components in PV systems[9].A micro-inverter and a aging failures.Ref.[57]investigates different circuit topologies of the single-phase PV inverters.Results indicate that failures often occur in switching stage and temperature is the most likely cause of failure.

(1)Failure rates of power electronic switches

The empirical formula of calculating the failure rate of IGBT and MOSFET can be found in[58,59]respectively.It is observed that the failure rates of IGBT and MOSFET are largely determined by ther-mal overstress[60].That means the failure rates of switches are re-lated to power losses and system power input levels since the failure rate is the functions of voltage or temperature,whereas the temperature depends on the power loss which in turn relies on system power input levels.Meanwhile,the empirical formula of the failure rate of diode is given in Refs.[59,61].As diodes are af?liated to IGBT and MOSFET in the same case,the reliability of diodes is also dependent on power losses[62]and system power input levels through temperature and voltage.

Given the power losses,the steady-state temperature rise in

Centralized PV inverter(Photo courtesy of SMA Solar Technology).IGBT module for high-voltage,high-current PV System(Photo courtesy Technologies).

826P.Zhang et al./Applied Energy104(2013)822–833

that the most urgent problem affecting inverter reliability is the quality of the dc-bus capacitors’’[68].

Classical method to predict reliability of electrolytic capacitors can be found in MIL-HDBK-217,in which the failure rate of a capacitor is dependent on the applied DC voltage,ripple current and ambient conditions(temperature,air?ow,and heat sinking). In particular,PV systems mounted outdoor may suffer from a rel-atively high failure rate of capacitors due to the harsher ambient environment.The failure rate of capacitor is,therefore,mainly determined by the core temperature,which can be calculated by the base life at elevated maximum core temperature and the actual core temperature[69].This failure rate formula is derived from the Arrhenius’s law[70],and in agreement with the‘‘life doubles every 10°C’’rule for capacitors.

Capacitor impedance can be modeled as an equivalent circuit in Fig.7where C R is the ideal capacitance,ESR is the ideal internal resistance,R S is the equivalent series resistance of the lead and junction resistance and ESL is the inductance of the electrodes. Among the above four parameters,C R and R S are the dominating temperature-dependent parameters.It was found that R S has a negative temperature dependency while C R has a positive temper-ature dependency.In other words,the equivalent series resistance decreases and the ideal capacitance increases as the temperature rises[71].The core temperature can be estimated by the superpo-sition of the ambient temperature and the temperature rise caused by ripple current[72].For the single-phase inverter shown in Fig.4,an approximate ripple current formula with an assumption of no energy storage in the inverter can be found in[73].For other inverters with more complex con?gurations,however,the current ripple through capacitor is almost caused by the switching of the switches and highly relies on the switching pattern[74].The read-ers who are interested in this issue can refer to[75–78]for more knowledge.

(3)Inverter topologies

Besides component reliability analysis for inverters,some work has been conducted on the reliability of various structures of inverters.For instance,the reliability of a single-stage three-phase integrated inverter is investigated in[11],where the thermal behavior is integrated into the reliability model of PV system.In [22],the reliability of more inverter con?gurations is studied, including an integrated topology,a two-stage con?guration,and a three-stage one.Different connections between the modules and inverters,i.e.,the AC-bus level and DC-bus level connections are explored in[41].Results show that higher system reliability can be achieved by using module-integrated inverters.A system-atic approach to studying the reliability of power-electronic com-ponents in a PV inverter can be found in[7].Ref.[79]presents a coherent methodology for integrating reliability considerations into the design of fault-tolerant power electronic systems and Ref.[80]proposes an optimal design methodology for PV inverters without transformer.It calculates the optimal con?guration of components by minimizing the leveraged cost of electricity(LCOE)which takes into consideration of the failure rate of components. This optimal design methodology may help lower the manufactur-ing and maintenance costs of the PV converters.

3.2.Reliability evaluation methods of PV system

As shown in Figs.1–3,a PV system consists of n PV strings.Each string is responsible for one-n th power output of the entire PV sys-tem,which means that the failure of some PV strings will not lead to the failure of the whole PV system but will only decrease the PV output.This is a key idea in reliability evaluation of PV system.

(1)Markov process method

The stochastic behavior of a PV system can be viewed as a Mar-kov process and described by a Markov space state diagram.The main advantage of this method is the clear picture of all states and transitions between the states.The transitions between vari-ous Markov states are due to failures and repairs of PV strings/ modules/inverters.Hence,this method can be very useful for mod-eling the outages of individual components.By solving the state transition matrix,the steady states of the Markov model can be ob-tained.The primary outputs of the Markov model are the steady-state probabilities and the duration residing in each state.Based on the Markov method,economic costs due to component failures can be calculated[81]by introducing cost rates to each state and cost impulses to the transitions in the Markov chain.The Markov-ian framework proposed in[25]provides performance-related metrics(e.g.energy yield)on top of the traditional reliability mod-els(e.g.MTBF).However,the Markov chain method suffers the curse of dimensionality and its application is restricted to low-dimensional spaces.Additionally,this method did not address the intermittent nature of solar inputs.Thus,more researches are required to reformulate the reward vector to introduce input uncertainty to the PV energy yield estimation.Meanwhile,Markov state space diagrams are drawn in[14,25]for reliability evaluation of the central-inverter PV system and distributed-inverter PV system.

(2)Monte-Carlo simulation

As an often-used method,Monte-Carlo simulation is also used in reliability analysis of PV systems[7,16,82].For a highly complex system,Monte-Carlo simulation is preferred because its computa-tional ef?ciency is independent of either the size or complexity of the system.Conceptually,Monte-Carlo simulation owns much more?exibility and can be used to study more complicated prob-lems,such as reliability assessment of PV-Wind hybrid system [83,84].There are two types of Monte-Carlo(MC)simulation: sequential MC and non-sequential MC.Sequential MC creates sys-tem states based on transitional probabilities or state durations and the correlations between the chronologically-sampled random variables can be included[85,86].It has been used to quantify the reliability indices of microgrid consisting of both wind and PV gen-erations.Results show that this type of microgrid is more unreli-able than the systems with conventional generation[85].On the other hand,non-sequential MC samples states based on their prob-abilities[87].The main advantages of sequential MC are that it can provide(1)more accurate evaluation of frequency and duration indices;(2)detailed modeling of state-duration transitions and (3)the ability to calculate the probability distributions of the risk indices.But sequential MC often requires much longer time to reach convergence than non-sequential MC.Pseudo-chronological MC simulation was proposed to retain the ef?ciency of non-sequential MC and to model chronological loads in sequential MC [88].This method was demonstrated on the IEEE-MRTS

(Modi?ed P.Zhang et al./Applied Energy104(2013)822–833827

Reliability Test System)[89],Results show that it took the compu-tational effort similar to that required by non-sequential MC,but with much better accuracy for chronological load patterns.

(3)State Enumeration

State Enumeration Method(SEM)is used in[90,91]to compute the reliability indices of PV system.This method accounts for the impacts of power inputs,voltage levels and power losses on the failure rates of panel components.Each PV string has two mutually exclusive states:the working state and the out-of-service state. First,the equivalent reliability parameters for all PV strings in a PV array are computed.Then,the probability of an enumerated state of the PV array can be obtained by multiplying the equivalent availability of the functional strings and the equivalent unavail-ability of the failed strings.After that,all enumerated states with the speci?c number of failed strings are aggregated into the spe-ci?c state of the PV array.Finally,the state probabilities obtained using SEM are combined with other system parameters to deter-mine reliability indices of the PV system.There are two types of indices:time-oriented and energy-oriented.More details about the indices are given in Section3.3.Furthermore,readers are rec-ommended to refer to Ref.[90]for more details about SEM.As a generic and?exible method,SEM is applicable to any structure such as centralized-/string-/micro-inverter structure,and also to both the homogenous and heterogeneous PV strings.

(3)Reliability Block Diagram

Using Reliability Block Diagram(RBD),Ref.[15]develops the Photovoltaic Reliability and Performance Model(PV-RPM).The combined model can predict PV system energy output when taking into account the availability of components,solar irradiance,and module and inverter performance.PV-RPM consists of three com-ponents:Failure modes and effects analysis(FMEA),accelerated life tests and system reliability/availability modeling.FMEA helps systematically identify,analyze and document all the possible fail-ure scenarios and their impacts on the rest of the system.Acceler-ated life test runs the components,such as PV panels,under elevated stress to collect the time-to-failure data.System reliabil-ity/availability model is a diagram that represents all subsystem and component events that must occur for successful system oper-ation.Recently,failure modes,effects and criticality analysis(FME-CA)for PV system is proposed to provide an understanding of the system failure modes[92].The criticality in the FMECA is a quan-titative index scale that represents the seriousness of failure modes.This enables us to do priority ranking among all the failure modes and their impacts on the system.However,FMECA is an inductive analysis method which requires a profound and detailed knowledge about every single failure mode.

(5)Fault Tree Analysis

Fault Tree Analysis(FTA)was?rst developed in1961by the US Air Force.It translates a physical system into a structured logic dia-gram known as fault trees.It not only considers the basic events that cause failures,but also represents various logical relationships associated with ambient conditions and human errors causing fail-ures.In general,FTA consists of four basic steps[93,94]:(1)system de?nition;(2)fault-tree construction;(3)qualitative evaluation; and(4)quantitative evaluation.

Although FTA is a powerful tool for reliability assessment,it re-quires considerable non-computerized efforts in constructing fault trees.Ref.[95]analyzes simple stand-alone PV systems using the Failure Mode Effect Analysis(FMEA)and FTA.Three typical solar photovoltaic systems are discussed in this paper.In[96],a method based on FTA is proposed for assessing the reliability of large-scale grid-connected photovoltaic systems.In[9],FTA and Markov pro-cess method are jointly used to describe the behavior of PV system. The life-cycle energy cost of PV system is calculated and applied to PV system designs.

3.3.Reliability indices for PV system

Reliability indices for traditional power distribution systems have been widely accepted by power industry and regulation bodies,and have been documented as international standards [97–100].For PV systems,these reliability indices are also helpful in selecting the best design in the planning stage and in reducing cost and increasing bene?t in operational stage.The traditional reliability indices such as mean time between failures(MTBFs), mean time to repairs(MTTRs),loss of load probability(LOLP), and loss of load hours(LOLHs)have been adopted in the literatures [7,101,102].Ref.[103]introduced the loss of power probability (LPP)index which considered the extreme values of data as func-tions of certain recurrence intervals.However,the index is limited to stand-alone PV system and a constant daily load was unrealisti-cally assumed in the model.

On the other hand,because of the fault tolerant design,a PV sys-tem may remain operational at a reduced power output level. Moreover,the maintenance of PV systems can be conducted at night with little or no effect on system availability.The traditional reliability indices may not completely satisfy the requirements of PV system.Some new reliability indices are developed for the reli-ability evaluation of PV system.

The Yearly Expected Energy Production(YEEP)index is used to evaluate PV system reliability in[104].The YEEP is obtained based on a multi-state system model by considering both component failures and PV power outputs.Additionally,the expected energy production can be calculated by the universal generating function (UCF)technique which is based on the State Enumeration ap-proach.Similarly in[91],the states of PV system are further dis-tributed into the three subsets(full available,derated,and outage states)according to the various PV power outputs.At the same time,two kinds of new reliability indices,called by the en-ergy-oriented index and time-oriented index,are de?ned to quan-tify the PV system reliability.The former index describes the energy availability and is de?ned as the expected output energy normalized by the ideal output energy,whereas the latter one de-scribes the time availability and is determined by the available, derated and outage hours.Note that the energy availability is more yield oriented,whereas the time availability is more cost oriented due to repairs and outages.Result shows that the energy availabil-ity index can indicate both the PV degradation and aging failure while the time availability index mainly re?ects the impact of aging failures.Ref.[105]discussed subdivided state sets in calcu-lating the two indices.

4.Future perspectives on PV reliability assessment

4.1.Cyber-physical system perspective on reliability assessment of PV system

A PV power system was treated as a pure physical system in existing reliability assessment techniques.A real-life PV system, however,is a cyber-physical system,where the cyber part plays a critical role in system operations.Major cyber components in-clude all power conversion algorithms,inverter state machine for different operating modes,maximum power point tracker(MPPT), digital phase-locked-loop(PLL),islanding detector and fault pro-tection.Failures or malfunctioning of these software and control

828P.Zhang et al./Applied Energy104(2013)822–833

components will reduce energy output or even cause frequent shutdown or outage in PV systems.Therefore,the reliability of the software and control needs to be modeled and incorporated into overall PV system reliability.

The fault injection method[106,107]will be a promising tech-nology to quantify the cyber reliability of PV systems.First,a benchmark PV system[108](see Fig.8)will be connected to a real-time hardware-in-the-loop simulator through signal condi-tioning interface.Then,the real-time simulator will randomly in-ject erroneous signals or faults to digital controllers following a typical software model such as logarithmic–exponential growth model,and PV system responses will be checked.A PV system will run into error state if there is any violation of performance bounds such as grid overvoltage/overvoltage,grid frequency excursion, switch overcurrent,peak inverter output current,and PV overvolt-age/undervoltage.Finally,the failure modes in the cyber part and corresponding consequences will be statistically obtained.

4.2.Modeling voltage control scheme in reliability assessment

Component failure is not the only reason that causes shutdown of grid-connected PV power system.Overvoltage caused by reverse power?ow is another major factor that limits the power output from PV power systems[109].This causes transient voltage?uctu-ation and may increase harmonic distortion of the network voltage [110]and lead to poor voltage quality.The current industrial prac-tice is that,when PV terminal voltage exceeds a pre-speci?ed va-lue,overvoltage protection in PV inverter will directly shut down the PV system[111].Due to the highly intermittent nature of solar power,traditional overvoltage protection will inevitably cause fre-quent interruptions of PV power supply,leading to signi?cant reduction of total PV energy output,loss of revenue,poor power quality,and higher risks due to deterioration of aging equipment in distribution grid.

Direct active power control of PV power generation has become increasingly attractive because of its effectiveness and ef?ciency. Physically,a distinguishing feature of distribution network is the high R/X ratio in the network,and thus the voltage pro?le in distri-bution network is sensitive to active power variations in it.There-fore,fast control of PV active power output is an effective and economical measure to mitigate the voltage rise problem in distri-bution grid.Since voltage regulation scheme will be widely imple-mented in PV power systems,a possible research is to investigate how to incorporate an online active power control[112–114]into PV reliability assessment.

Recently,Ref.[115]compared the dynamic characteristics of centralized and distributed(micro-inverter)PV generations using both conventional power factor control and Generator Emulator Control(GEC)for PV inverters.It was found that the GEC method may mitigate voltage rise of both centralized and distributed PV systems by utilizing shunt capacitors and line voltage regulators. Additionally,GEC method may compensate reactive power based on the slope of droop curves.Hence,another possible research is to investigate how to incorporate GEC into PV reliability assessment.

Voltage control scheme will change the input power of PV in-verter and thus power losses in power electronics switches and ripple current through capacitors.These changes will affect the failure rates of electronics components through temperature vari-ations.Therefore,the control actions will signi?cantly affect the reliability of PV power system.A sequential simulation approach may be adopted to incorporate voltage control effect on PV reliabil-ity.Basic procedures may include the following:

(1)Read chronological solar power records and parameters of

distribution network to which a PV system is connected.

(2)Calculate adaptive P-V droops.

(3)Obtain the chronological power output of PV inverter with

the PV voltage control scheme.

(4)Use the inverter power curve obtained as the data input for

PV reliability assessment.

4.3.PV reliability assessment under extreme weather conditions

Distributed PV generation can mitigate the damaging effects of extreme weather conditions such as hurricanes,earthquakes,tsu-nami,tropical storms,and snow storm.For instance,Hurricane ISAAC hit the state of Louisiana,USA in August,2012and left more than200,000residents without power for weeks;a historic earth-quake and tsunami hit Fukushima,Japan in March,2011and led to the multiple catastrophic nuclear reactor meltdown disaster and left millions of people without power for months.All these ex-treme events caused serious prolonged power outages,resulting in millions of dollars to restore power and billions of dollars of eco-nomic losses.This is why more researches are required to improve resiliency and reliability of distribution system with consideration of extreme events.

Ref.[116]proposed a sequential Monte-Carlo simulation for PV system with consideration of adverse weather effects.First,weath-er conditions are classi?ed as either normal or adverse.The deci-sion criterion depends on its impact on the failure rate of system components.‘‘Normal’’weather conditions should have little or no effect on component failure rates whereas‘‘adverse’’weather conditions will have a large impact on the failure rates.Second, the component failure rate considering the weather impact is the product of the average failure rate and the failure rate weather fac-tor.Similarly,the restoration time considering the weather impact can be obtained by the product of the restoration time weather

fac-Fig.8.Experimental setup of the benchmark PV system(a)Block diagram and(b)DSP-controlled power interface.

tor,average restoration time in normal condition and restoration time factor.Third,the PV generation with varying insolation inten-sity(G t)and weather-related failure and restoration rates is repre-sented by a three-state model as shown in Fig.9.Finally,the reliability indices of the PV-connected distribution system are cal-culated by the sequential Monte-Carlo simulation.It should be pointed out that the two condition expression of normal and ad-verse weather has been used in weather-related reliability evalua-tion for years and generally is not good for modeling extreme weather conditions.Modeling extreme weather conditions in reli-ability evaluation of PV system reliability or PV-connected distri-bution system is a real challenge.Multiple micro-grids with renewable energy sources and self-diagnosis of PV systems[117] may be considered as two measures to mitigate the impacts of ex-treme weather conditions in modeling.

4.4.PV reliability assessment considering cyber vulnerability,attack and security

As the grid becomes more reliable,intelligent,resilient,infra-structures may become a critical target by adversaries.If attacks were successful,they would cause serious damages to people’s lives,?nancial and social stabilities.It is obvious that cyber-vulner-ability should be taken into consideration in PV reliability assess-ment.This will not only help improve our understanding about the smart grid’s reliability more accurately,but also enable engi-neers and computer scientists to develop new security protocols to continuously build a more secured smart grid.Ref.[118]has identi?ed several types of possible attacks:(1)data integrity at-tack;(2)denial of service attack;(3)replay attacks;(4)timing at-tacks;and(5)de-synchronization attacks.

One potential method to improve the cyber security includes the use of the attack resilient control algorithm and domain-spe-ci?c anomaly detection and intrusion algorithm to protect the measurements and control signals on the communication plat-form.Further research is needed in this direction to implement these algorithms and test their effectiveness against security threats.Also,quanti?ed reliability assessment of the cyber-physi-cal system is needed.

5.Reliability evaluation of power grids with PV systems

5.1.Active distribution network including PV microgrids

A PV power system normally connects with local loads,other distributed generators(DGs),energy storages and protection and control devices,forming a microgrid interconnected to a local dis-tribution grid.To quantify the impact of PV interconnection on dis-tribution reliability,one needs to evaluate the reliability of distribution grid with embedded microgrids[66].

A PV microgrid can be either interconnected to distribution grid or autonomously operated as an island when detached from the main power supply due to faults or disturbances.As a result,to-day’s distribution systems are evolving from passive networks to active and smarter grids with various?exible operation modes. This brought two challenges to reliability evaluation of active dis-tribution systems.The?rst is how to model the intermittent and uncertain PVs.The second is the development of new evaluation methods for active distribution systems with embedded microgr-ids.Traditional distribution reliability evaluation models and methods,which were developed for passive radial networks,are no longer able to deal with modern active distribution network including distributed microgrids.This is because the methods can only deal with network with?xed power?ow directions.However, bidirectional power?ows caused by embedded microgrids and multiple contingencies have to be taken into consideration for ac-tive distribution network.To tackle the challenges,a new hybrid Monte Carlo method[119]has recently been proposed for reliabil-ity evaluation of active distribution network with microgrids.The method includes the following basic considerations:

(1)An active distribution network may consist of many PV

microgrids,each of which includes autonomous PV power generators and loads.With detailed representation of every component,the computational burden in the reliability eval-uation will become prohibitive.A virtual power plant(VPP) concept is used to aggregate all components in a PV micro-grid into a single entity.A VPP will be represented by a multi-state model in reliability analysis.

(2)A hybrid Monte Carlo simulation method is employed to

determine network states.A sequential simulation[120]is applied to simulate PV inverter power outputs and a non-sequential simulation is used to evaluate reliability indices of distribution system.

(3)For each sampled state,a combined minimal path and zone

partitioning technique is utilized to deal with state evalua-tion,which can handle islanding operation modes.

There are remaining challenges.First,in a real-life distribution system,there exist high correlations in solar insolation between PV generators if they are located in the close vicinity[121].The use of random number to determine an output for each PV gener-ator independently without considering the correlation between PVs could result in an optimistic reliability outcome of reliability indices,and hence could over-estimate the overall bene?t of PV installation.The second issue is the correlation between solar inso-lation and customer load level.For example,the PV output should be zero at night when the load is at off-peak[122].A random num-ber generation without considering the second correlation could also create simulation errors in PV reliability assessment results.

Nowadays,more and more smart-grid technologies,including demand response schemes[123],energy storages and electric vehicle charging stations[124],are being implemented in distribu-tion systems.These ubiquitous devices greatly alter the horizon of power system operations and planning[125].Their impacts on PV reliability and on the reliability of distribution system with microgrids should be carefully investigated.

Due to the multi-scale and temporal-spatial dynamics in mod-ern power system,new mathematical and engineering solutions have to be developed to evaluate the future PV-connected power system reliability.Sequential simulation method could be one of candidate

solutions. 830P.Zhang et al./Applied Energy104(2013)822–833

5.2.Reliability for future Giga-PV system connected to power transmission grid

As PV is becoming a popular source of renewable energy,inno-vative researchers from the US Defense Advanced Research Pro-jects Agency(DARPA)is trying to create superef?cient and compact PV panels that would convert up to50%solar energy into DC electricity[126].The heart of this research is to con?gure PV modules by the new‘‘side-by-side array’’design instead of the cur-rent‘‘multi-junction stacking’’design.Hence,with the promising superef?cient PV module development,it is obvious that the aver-age power output of PV plants may increase dramatically into mega-watt or even giga-watt capacity in the future.At a certain point,these Giga-PV plants,which may be directly interconnected to the grid at the transmission level rather than via feeders at dis-tribution voltage level,will become a reality[127].On the other hand,these large-scale Giga-PV plants will tend to be installed in remote areas such as the deserts because they take up more land compared with the output-equivalent wind farms or fuel cells. Due to the long distance between deserts and cities,high-voltage transmission lines are usually preferred to ef?ciently transport the electricity.However,there exist some challenges in terms of power quality,voltage,angular and frequency stability,etc. [128].The major cause of these challenges comes from the fact that inverters’power electronics often introduce harmonics and from the high intermittence of PV power output that may cause grid instability.This problem is ampli?ed as the total ratio of PV gener-ation increases.Thus,proper control strategies must be developed to ensure the grid stability in the future.

One possible research is to study the reliability of transmission integrated with high-voltage PV system,with considerations for power quality,energy adequacy,voltage and transient stability, as well as cyber reliability.Also,more researches on PV system reliability degradation due to exposure to harsh environments are needed.This is partly because as the high-voltage inverters are located in the hot desert,the failure rates of power electronics switches may be very high.

6.Conclusions

This paper reviews the methods for evaluating the reliability of large-scale PV systems and techniques for quantifying the effects of PV interconnection on distribution system reliability.It provides a survey of practical approaches to reliability analysis of PV invert-ers,PV modules and array,and overall PV power systems.The reli-ability analysis methods of active distribution systems embedded with PV microgrids are also summarized.Some major directions for future researches have been identi?ed.These include incorpo-rating cyber reliability models in PV reliability assessment;inte-grating control and protection schemes into the PV reliability models;PV reliability assessment under extreme weather condi-tions and attack events:modeling demand response,storage,EV charging and other smart grid facilities in the reliability evaluation of active distribution network with PV microgrid;and reliability assessment for Giga-PV system connected to power transmission grid.

References

[1]Sawin J,Martinot E.Renewables2010global status report.Renewable Energy

Policy Network;2010[Report No.:1].

[2]Lall P.Tutorial:temperature as an input to microelectronics-reliability

models.IEEE Trans Rel1996;45(1):3–9.

[3]Roy A,Kedare SB,Bandyopadhyay S.Optimum sizing of wind-battery systems

incorporating resource uncertainty.Appl Energy2010;87(8):2712–27.

[4]Durisch W,Tille D,W?rz A,Plapp W.Characterisation of photovoltaic

generators.Appl Energy2000;65(1–4):273–84.

[5]Ho WS,Hashim H,Hassim MH,Muis ZA,Shamsuddin NLM.Design of

distributed energy system through electric system cascade analysis(ESCA).

Appl Energy2012;99(0):309–15.

[6]Landgrebe AR,Donley SW.Battery storage in residential applications of

energy from photovoltaic sources.Appl Energy1983;15(2):127–37.

[7]Ristow A,Begovic M,Pregelj A,Rohatgi A.Development of a methodology for

improving photovoltaic inverter reliability.IEEE Trans Ind Electron 2008;55(7):2581–92.

[8]Roman E,Alonso R,Ibanez P,Elorduizapatarietxe S,Goitia D.Intelligent PV

module for grid-connected PV systems.IEEE Trans Ind Electron 2006;53(4):1066–73.

[9]Stember LH,Huss WR,Bridgman MS.A methodology for photovoltaic system

reliability and economic analysis.IEEE Trans Rel1982;R-31(3):296–303. [10]Smet V,Forest F,Huselstein J,Richardeau F,Khatir Z,Lefebvre S,et al.Ageing

and failure modes of IGBT modules in high-temperature power cycling.IEEE Trans Ind Electron2011;58(10):4931–4.

[11]Liu J,Henze N.In:Reliability consideration of low-power grid-tied inverter

for photovoltaic application.24th European photovoltaic solar energy conference and exhibition,September,Hamburg,Germany;2009.

[12]Petrone G,Spagnuolo G,Teodorescu R,Veerachary M,Vitelli M.Reliability

issues in photovoltaic power processing systems.IEEE Trans Ind Electron 2008;55(7):2569–80.

[13]Alonso R,Roman E,Sanz A,Martinez Santos VE,Ibanez P.Analysis of inverter-

voltage in?uence on distributed MPPT architecture performance.IEEE Trans Ind Electron2012;59(10):3900–7.

[14]Dhople SV,Davoudi A,Chapman PL,Domi′nguez-Garci′a AD.In:Integrating

photovoltaic inverter reliability into energy yield estimation with markov models.Control and modeling for power electronics(COMPEL),2010IEEE 12th workshop on;2010.p.1–5.

[15]Collins E,Dvorack M,Mahn J,Mundt M,Quintana M.In:Reliability and

availability analysis of a?elded photovoltaic system.Photovoltaic specialists conference(PVSC),200934th IEEE;2009.p.002316–21.

[16]Pregelj A,Begovic M,Rohatgi A.In:Impact of inverter con?guration on PV

system reliability and energy production.Photovoltaic specialists conference, 2002.Conference record of the twenty-ninth IEEE;2002.p.1388–91. [17]Meyer EL,van Dyk EE.Assessing the reliability and degradation of

photovoltaic module performance parameters.IEEE Trans Rel 2004;53(1):83–92.

[18]Shu Z,Jirutitijaroen https://www.wendangku.net/doc/dc5930730.html,tin hypercube sampling techniques for power

systems reliability analysis with renewable energy sources.IEEE Trans Power Syst2011;26(4):2066–73.

[19]Dhople SV,Davoudi A,Domi′nguez-Garci′a AD,Chapman PL.A uni?ed

approach to reliability assessment of multiphase DC–DC converters in photovoltaic energy conversion systems.IEEE Trans Power Electron 2012;27(2):739–51.

[20]Hirschmann D,Tissen D,Schroder S,de Doncker RW.In:Reliability prediction

for inverters in hybrid electrical vehicles.Power electronics specialists conference,2006.PESC‘06.37th IEEE;2006.p.1–6.

[21]Rodriguez C,Amaratunga GAJ.Long-lifetime power inverter for photovoltaic

AC modules.IEEE Trans Ind Electron2008;55(7):2593–601.

[22]Chan F,Calleja H.Reliability estimation of three single-phase topologies in

grid-connected PV systems.IEEE Trans Ind Electron2011;58(7):2683–9. [23]Kroposki B,Margolis R,Kuswa G,Torres J,Bower W,Key T,et al.Renewable

systems interconnection:executive summary.NREL Technical Report;

February2008[Report No.:NREL/TP-581-42292].

[24]Bebic J,Kroposki B.Power system planning:emerging practices suitable for

evaluating the impact of high-penetration photovoltaics.NREL Technical Report;February2008[Report No.:NREL/SR-581-42297].

[25]Dhople SV,Dominguez-Garcia AD.Estimation of photovoltaic system

reliability and performance metrics.IEEE Transactions Power Syst 2012;27(1):554–63.

[26]Gaikwad D,Mehraeen S.In:Reactive power considerations in reliability

analysis of photovoltaic systems.Green technologies conference,2012IEEE;

2012.p.1–6.

[27]Bae I,Kim J.Reliability evaluation of customers in a microgrid.IEEE

Transactions Power Syst2008;23(3):1416–22.

[28]Basu AK,Chowdhury S,Chowdhury SP.Impact of strategic deployment of

CHP-based DERs on microgrid reliability.IEEE Transactions Power Del 2010;25(3):1697–705.

[29]Patra SB,Mitra J,Ranade SJ.In:Microgrid architecture:a reliability

constrained approach,vol.3.Power engineering society general meeting, 2005.IEEE;2005.p.2372–7.

[30]Mitra J,Patra SB,Ranade SJ.In:Reliability stipulated microgrid architecture

using particle swarm optimization.Probabilistic methods applied to power systems,2006.PMAPS2006.international conference on;2006.p.1–7. [31]Basu AK,Chowdhury SP,Chowdhury S,Ray D,Crossley PA.In:Reliability

study of a micro-grid power system.Universities power engineering conference,2008.UPEC2008.43rd,international;2008.p.1–4.

[32]Nosrat A,Pearce JM.Dispatch strategy and model for hybrid photovoltaic and

trigeneration power systems.Appl Energy2011;88(9):3270–6.

[33]Kwasinski A.Quantitative evaluation of DC microgrids availability:effects of

system architecture and converter topology design choices.IEEE Trans Power Electron2011;26(3):835–51.

[34]Milo A,Gaztanaga H,Etxeberria-Otadui I,Bilbao E,Rodriguez P.In:

Optimization of an experimental hybrid microgrid operation:reliability and economic issues.PowerTech,2009IEEE bucharest;2009.p.1–6.

P.Zhang et al./Applied Energy104(2013)822–833831

[35]Karki R,Billinton R.Reliability/cost implications of PV and wind energy

utilization in small isolated power systems.IEEE Trans Energy Convers 2001;16(4):368–73.

[36]Li Z,Yuan Y,Li F.In:Evaluating the reliability of islanded microgrid in an

emergency mode.Universities power engineering conference(UPEC),2010 45th,international;;2010.p.1–5.

[37]Park J,Wu L,Choi J,El-Keib AA,Shahidehpour M,Billinton R.In:A

probabilistic reliability evaluation of a power system including solar/ photovoltaic cell generator.Power&energy society general meeting,2009.

PES09.IEEE2009.p.1–6.

[38]Fotuhi-Firuzabad M,Rajabi-Ghahnavie A.In:An analytical method to

consider DG impacts on distribution system reliability.Transmission and distribution conference and exhibition:Asia and paci?c,2005IEEE/PES;2005.

p.1–6.

[39]Sun Y,Bollen MHJ,Ault GW.In:Probabilistic reliability evaluation for

distribution systems with DER and microgrids.Probabilistic methods applied to power systems,2006.PMAPS2006.International conference on;2006.p.

1–8.

[40]Celli G,Ghiani E,Soma GG,Pilo F.In:Active distribution network reliability

assessment with a pseudo sequential Monte Carlo method.PowerTech,2011 IEEE trondheim;2011.p.1–8.

[41]Rohouma WM,Molokhia IM,Esuri https://www.wendangku.net/doc/dc5930730.html,parative study of different PV

modules con?guration reliability.Desalination2007;209:122–8.

[42]Carbone R,Tomaselli A.In:Recent advances on AC PV-modules for grid-

connected photovoltaic plants.Clean electrical power(ICCEP),2011 international conference on;2011.p.124–9.

[43]Sánchez Reinoso CR,Milone DH,Buitrago RH.Simulation of photovoltaic

centrals with dynamic shading.Appl Energy,in press.

[44]Kim J,Jeon G,Hong W.The performance and economical analysis of grid-

connected photovoltaic systems in daegu,korea.Appl Energy 2009;86(2):265–72.

[45]Ould-Amrouche S,Rekioua D,Hamidat A.Modelling photovoltaic water

pumping systems and evaluation of their CO2emissions mitigation potential.

Appl Energy2010;87(11):3451–9.

[46]Kumar R,Rosen MA.A critical review of photovoltaic–thermal solar collectors

for air heating.Appl Energy2011;88(11):3603–14.

[47]Chow TT.A review on photovoltaic/thermal hybrid solar technology.Appl

Energy2010;87(2):365–3679.

[48]Orioli A,Di Gangi A.A procedure to calculate the?ve-parameter model of

crystalline silicon photovoltaic modules on the basis of the tabular performance data.Appl Energy,in press.

[49]Sharma R,Tiwari GN.Technical performance evaluation of stand-alone

photovoltaic array for outdoor?eld conditions of new delhi.Appl Energy 2012;92(0):644–52.

[50]Dumas LN,Shumka A.Photovoltaic module reliability improvement

through application testing and failure analysis.IEEE Trans Rel1982;R-31(3):228–34.

[51]Wohlgemuth JH,Cunningham DW,Nguyen AM,Miller J.In:Long term

reliability of PV module.20th European photovoltaic solar energy conference, Barcelona,Spain;2005.p.1942–6.

[52]Catelani M,Ciani L,Cristaldi L,Faifer M,Lazzaroni M,Rossi M.In:

Characterization of photovoltaic panels:the effects of dust.Instrumentation and measurement technology conference(I2MTC),2012IEEE,international;

2012.p.1–4.

[53]Vazquez M,Rey-Stolle I.Photovoltaic module reliability model based on?eld

degradation studies.Prog Photovoltaics:Res Appl2008August;16(5):419–33.

[54]Klein SA,Beckman WA.Loss-of-load probabilities for stand-alone

photovoltaic systems.Solar Energy1987;39(6):499–512.

[55]Ross RG.In:Photovoltaic module and array reliability.15th IEEE photovoltaic

specialists’conference;Orlando,Florida;1981.p.499–512.

[56]Gautam NK,Kaushika ND.Reliability evaluation of solar photovoltaic arrays.

Solar Energy2002February;72(2):129–41.

[57]Chan F,Calleja H,Martinez E.In:Grid connected PV systems:a reliability-

based comparison.Industrial electronics,2006IEEE international symposium on;2006.p.1583–8.

[58]FIDES Guide2009.Reliability methodology for electronics systems.FIDES

Group;2009.

[59]Military Handbook MIL-HDBK-217Notice 2.Reliability prediction of

electronic equipment.Washington(DC):US Department of Defense;1995.

[60]Tiwari GN,Mishra RK,Solanki SC.Photovoltaic modules and their

applications:a review on thermal modelling.Appl Energy 2011;88(7):2287–304.

[61]Military Handbook MIL-HDBK-217F.Reliability prediction of electronic

equipment.Washington(DC):US Department of Defense;1991.

[62]Arifujjaman M,Iqbal MT,Quaicoe JE.Reliability analysis of grid connected

small wind turbine power electronics.Appl Energy2009;86(9):1617–23. [63]Luo Z,Ahn H,Nokali MAE.A thermal model for insulated gate bipolar

transistor module.IEEE Trans Power Electron2004;19(4):902–7.

[64]Chow TT,Chan ALS,Fong KF,Lin Z,He W,Ji J.Annual performance of building-

integrated photovoltaic/water-heating system for warm climate application.

Appl Energy2009;86(5):689–96.

[65]Nelson JJ,Venkataramanan G,El-Refaie AM.Fast thermal pro?ling of power

semiconductor devices using fourier techniques.IEEE Trans Ind Electron 2006;53(2):521–9.

[66]Wei L,Kerkman RJ,Lukaszewski RA,Brown BP,Gollhardt N,Weiss BW.In:

Junction temperature prediction of a multiple-chip IGBT module under DC

condition.Industry applications conference,2006.41st IAS annual meeting.

conference record of the2006IEEE;2006.p.754–62.

[67]Harb S,Balog RS.In:Reliability of candidate photovoltaic module-integrated-

inverter topologies.Applied power electronics conference and exposition (APEC),2012twenty-seventh annual IEEE;2012.p.898–903.

[68]Ton D,Bower W.Summary report on the DOE high-tech inverter workshop.

https://www.wendangku.net/doc/dc5930730.html, Department of Energy;2005.

[69]Parler SG.Deriving life multipliers for electrolytic capacitors.IEEE Power

Electron Soc Newslett2004;16(1):11–2.

[70]Albertini A,Masi MG,Mazzanti G,Peretto L,Tinarelli R.Toward a BITE for

real-time life estimation of capacitors subjected to thermal stress.IEEE Trans Instrum Meas2011;60(5):1674–81.

[71]Graditi G,Adinol?G.In:Temperature in?uence on photovoltaic power

optimizer components reliability.Power electronics,electrical drives, automation and motion(SPEEDAM),2012International symposium on;

2012.p.1113–8.

[72]Application guide,aluminum electrolytic capacitors.Liberty,SC:Cornell

Dubilier;2003.

[73]Rashid M.Power electronics handbook.2010:1409.

[74]Hirschmann D,Tissen D,Schroder S,De Doncker RW.In:Inverter design for

hybrid electrical vehicles considering mission pro?les.Vehicle power and propulsion,2005IEEE conference;2005.p.6

[75]Evans PD,Hill-Cottingham RJ.DC link current in PWM inverters.Electr Power

Appl,IEE Proc B1986;133(4):217–24.

[76]Mariscotti A.Analysis of the DC-link current spectrum in voltage source

inverters.IEEE Trans Circ Syst I:Fund Theory Appl2002;49(4):484–91. [77]Shah R,Mithulananthan N,Bansal RC.Damping performance analysis of

battery energy storage system,ultracapacitor and shunt capacitor with large-scale photovoltaic plants.Appl Energy2012;96(0):235–44.

[78]Kolar JW,Round SD.Analytical calculation of the RMS current stress on the

DC-link capacitor of voltage-PWM converter systems.Electr Power Appl,IEE Proc2006;153(4):535–43.

[79]Dominguez-Garcia AD,Krein PT.In:Integrating reliability into the design of

fault-tolerant power electronics systems.Power electronics specialists conference,2008.PESC2008.IEEE2008.p.2665–71.

[80]Koutroulis E,Blaabjerg F.Design optimization of transformerless grid-

connected PV inverters including reliability.IEEE Trans Power Electron 2013;28(1):325–35.

[81]Anders GJ,Leite da Silva AM.Cost related reliability measures for power

system equipment.IEEE Trans Power Syst2000;15(2):654–60.

[82]Maish AB,Atcitty C,Hester S,Greenberg D,Osborn D,Collier D,et al.In:

Photovoltaic system reliability.Photovoltaic specialists conference,1997, conference record of the twenty-sixth IEEE;1997.p.1049–54.

[83]Kishore LN,Fernandez E.In:Reliability well-being assessment of PV-wind

hybrid system using monte carlo simulation.Emerging trends in electrical and computer technology(ICETECT),2011International conference on;2011.

p.63–8.

[84]Saheb-Koussa D,Haddadi M,Belhamel M.Economic and technical study of a

hybrid system(wind–photovoltaic–diesel)for rural electri?cation in algeria.

Appl Energy2009;86(7–8):1024–30.

[85]Ghahderijani MM,Barakati SM,Tavakoli S.In:Reliability evaluation of stand-

alone hybrid microgrid using sequential monte carlo simulation.Renewable energy and distributed generation(ICREDG),2012Second iranian conference on;2012.p.33–8.

[86]Billinton R,Hua Chen,Ghajar R.A sequential simulation technique for

adequacy evaluation of generating systems including wind energy.IEEE Trans Energy Convers1996;11(4):728–34.

[87]Vallee F,Lobry J,Deblecker O.Impact of the wind geographical correlation

level for reliability studies.IEEE Trans Power Syst2007;22(4):2232–9. [88]Leite da Silva AM,Da Fonseca Manso LA,De Oliveira Mello JC,Billinton R.

Pseudo-chronological simulation for composite reliability analysis with time varying loads.IEEE Trans Power Syst2000;15(1):73–80.

[89]Subcommittee PM.IEEE reliability test system.IEEE Trans Power Apparatus

Syst1979;PAS-98(6):2047–54.

[90]Zhang P,Wang Y,Xiao W,Li W.Reliability evaluation of grid-connected

photovoltaic power systems.IEEE Trans Sust Energy2012;3(3):379–89. [91]Wang Y,Zhang P,Li W.In:Comparative analysis of the reliability of grid-

connected photovoltaic power systems.IEEE PES general meeting;San Diego;

2012.

[92]Catelani M,Ciani L,Cristaldi L,Faifer M,Lazzaroni M,Rinaldi P.In:FMECA

technique on photovoltaic module.Instrumentation and measurement technology conference(I2MTC),2011IEEE;2011.p.1–6.

[93]Lee WS,Grosh DL,Tillman FA,Lie CH.Fault tree analysis,methods,and

applications:a review.IEEE Trans Rel1985;R-34(3):194–203.

[94]Hamdy MA,Beshir ME,Elmasry SE.Reliability analysis of photovoltaic

systems.Appl Energy1989;33(4):253–6.

[95]Stember LH.Reliability considerations in the design of solar photovoltaic

power systems.Solar Cells1981;3(3):269–85.

[96]Zini G,Mangeant C,Merten J.Reliability of large-scale grid-connected

photovoltaic systems.Renew Energy2011;36(9):2334–40.

[97]Li W.Risk assessment of power systems.Wiley-IEEE;2005.

[98]IEEE guide for electric power distribution reliability indices.IEEE Std1366–

2003(Revision of IEEE Std1366–1998);2004:0–1.

[99]Desideri U,Proietti S,Zepparelli F,Sdringola P,Bini S.Life cycle assessment of

a ground-mounted1778kWp photovoltaic plant and comparison with

traditional energy production systems.Appl Energy2012;97(0):930–43.

832P.Zhang et al./Applied Energy104(2013)822–833

[100]Lauffenburger HA,Anderson RT.Reliability terminology and formulae for photovoltaic power systems.IEEE Trans Rel1982;R-31(3):289–95.

[101]Zhou W,Lou C,Li Z,Lu L,Yang H.Current status of research on optimum sizing of stand-alone hybrid solar–wind power generation systems.Appl Energy2010;87(2):380–9.

[102]Chen H.Optimum capacity determination of stand-alone hybrid generation system considering cost and reliability.Appl Energy,in press.

[103]Chen S.An ef?cient sizing method for a stand-alone PV system in terms of the observed block extremes.Appl Energy2012;91(1):375–84.

[104]Ding Y,Ostergaard J,Shen W,Wang P,Loh PC.In:Expected energy production evaluation for photovoltaic systems.Electric utility deregulation and restructuring and power technologies(DRPT),20114th International conference on;2011.p.1076–80.

[105]Warren C,Saint B.In:IEEE reliability indices standards-major event day calculations and how it effects small utilities.Rural electric power conference,2004;2004.p.D2,1–12.

[106]Hsueh Mei-Chen,Tsai TK,Iyer RK.Fault injection techniques and tools.

Computer1997;30(4):75–82.

[107]Natella R,Cotroneo D,Duraes J,Madeira H.On fault representativeness of software fault injection.Software Engineering,IEEE Transactions on 2012;PP(99):1.

[108]Xiao W,Zhang P.Photovoltaic voltage regulation using af?ne parameterization.Green Energy,in press.

[109]Masters CL.Voltage rise:The big issue when connecting embedded generation to long11kV overhead lines.Power Eng J2002;16(1):5–12. [110]Lakervi E,Holmes EJ,Institution of electrical engineers.Electricity distribution network design.2nd ed.London:P.Peregrinus on behalf of the Institution of Electrical Engineers;1995.

[111]Ueda Y,Kurokawa K,Tanabe T,Kitamura K,Sugihara H.Analysis results of output power loss due to the grid voltage rise in grid-connected photovoltaic power generation systems.IEEE Trans Ind Electron2008;55(7):2744–51. [112]Wang Y,Zhang P,Li W.Online overvoltage prevention control of photovoltaic generators in microgrids.IEEE Trans of Smart Grid,in press.

[113]Zhang P,Xue H,Yang R.Shifting window average method for accurate frequency measurement in power systems.IEEE Trans Power Del 2011;26(4):2887–9.

[114]Tonkoski R,Lopes LAC,El-Fouly THM.Coordinated active power curtailment of grid connected PV inverters for overvoltage prevention.IEEE Trans Sust Energy2011;2(2):139–47.[115]Gudimetla B,Katiraei F,Aguero JR,Enslin JHR,Alatrash H.In:Integration of micro-scale photovoltaic distributed generation on power distribution systems:dynamic analyses.Transmission and distribution conference and exposition(T&D),2012IEEE PES;2012.p.1–7.

[116]Cha ST,Jeon DH,Bae IS,Lee IR,Kim JO.In:Reliability evaluation of distribution system connected photovoltaic generation considering weather effects.Probabilistic methods applied to power systems,2004international conference on;2004.p.451–6.

[117]Kim GS,Lee K.In:Fault diagnosis and fault-tolerant control of a dc-link voltage sensor for PV inverters.Power electronics and motion control conference(IPEMC),20127th,international;2012.p.1408–12.

[118]Sridhar S,Hahn A,Govindarasu M.In:Cyber attack-resilient control for smart grid.Innovative smart grid technologies(ISGT),2012IEEE PES;2012.p.1–3. [119]Bie Z,Zhang P,Li G,Hua B,Meehan M,Wang X.Reliability evaluation of active distribution systems including microgrids.IEEE Trans Power Syst 2012;PP(99):1–9.

[120]Billinton R,Li W.Reliability assessment of electric power systems using monte carlo methods.New York:Plenum Press;1994.

[121]Zhang P,Li W.In:Reliability-oriented distribution network recon?guration.

Innovative smart grid technologies(ISGT),2010;2010.p.1–6.

[122]Zhang P,Li W.Boundary analysis of distribution reliability and economic assessment.IEEE Trans Power Syst2010;25(2):714–21.

[123]Liu Y,Zhang P.Optimal volt/var control in distribution systems.Int J Electric Power Energy Syst2002;24(4):271–6.

[124]Yu FR,Zhang P,Xiao W,Choudhury https://www.wendangku.net/doc/dc5930730.html,munication systems for grid integration of renewable energy https://www.wendangku.net/doc/dc5930730.html,work,IEEE2011;25(5): 22–9.

[125]Zhang P.C2E2photovoltaic power system:An advanced,scalable clean energy research platform for resolving energy challenges.Internal Report.

Center for Clean Energy Engineering(C2E2),University of Connecticut;2011. [126]Stevenson R.Tapping the power of100suns.IEEE Spectrum Magazine-Green Tech–Solar;2012.

[127]Walling RA,Clark K.In:Grid support functions implemented in utility-scale PV systems.Transmission and distribution conference and exposition,2010 IEEE PES;2010.p.1–5.

[128]Yang W,Zhou X,Xue F.In:Impacts of large scale and high voltage level photovoltaic penetration on the security and stability of power system.

Power and energy engineering conference(APPEEC),2010asia-paci?c;2010.

p.1–5.

P.Zhang et al./Applied Energy104(2013)822–833833

CTI华测检测可靠性实验室介绍

CTI Reliability Laboratory CTI Reliability Laboratory Introduction

Reliability Laboratory Machinery Surface Analysis Climate y C H L Mechani Mechan A Constant Te Humidity Dam Speedy Ther Solar Rad Gas Corr IP HA Cross Coating Tin Whiske Sold Cros Comp C Low / High cal Vibrati ical Shock Bump Drop Abrasion emperature mp Heat Temperat mal Shock diation \UV rosion\Sal ALT&HASS s Section g Thicknes er Observa derability ss-cut Tap ponent Ana Code h Temperat on k e and ture k V t Fog S ss ation e alysis ure

Speedy Temperature Change Test Model：ESPEC QW0470W10 Temperature Scope: -70℃-150℃ Change Speed：20℃/min Change Speed Humidity：25%RH～98%RH Equipment Size：700*750*700mm

Thermal Shock Test Model：ESPEC TSG0765W Temperature Scope: -65℃-150℃ T t S65150 Temperature Resume Time≤5min Equipment Size：410*460*370mm g g Bracket Max. Loading：30kg

三种可靠性试验比较

下文是对常见的三种可靠性试验进行分析和比较，可从试验目的、试验条件、试验方案或项目、试验合格性和受试产品的失效判据五个方面进行 (一)试验目的 A.可靠性增长试验在研制过程中模拟实际的或加速的使用条件进行试验，使产品存在的设计（包括电路设计、结构设计和工艺设计）缺陷变为硬故障而充分暴露，对故障进行分析、采取纠正措施，根除故障产生的原因或降低故障率到可以接受的值，使产品的固有可靠性得到增长。 B.可靠性鉴定试验验证产品的设计能否在规定的环境条件下满足规定的性能及可靠性要求。试验结果作为判断设备能否定型的依据。适用于设计定型的鉴定。 C.ORT 试验对产品各项指标进行全面检验，以评定产品质量和可靠性是否全部符合标准和达到设计要求。对于批量生产的产品检验其质量稳定性和一致性。适用于生产定型、批量生产后的一定周期和在产品设计、工艺、材料有较大变动后的检验。 (二)试验条件①电应力 A.可靠性增长试验根据输入交流电源电压和输入直流电源电压的允许变化范围，部分时间在设计的标称输入电压下工作，部分时间在最高输入电压下工作，部分时间在最低输入电压下工作。例如：程控用户交换机应在AC220V，DC-48V、DC-40V～-57V范围内正常完成接续。 B.可靠性鉴定试验同可靠性增长试验 C.ORT 试验除电源电压拉偏试验外，在标称输入电压下工作。电源拉偏试验根据不同的产品参考有关标准在最高、最低电压下工作。 (二)试验条件②热应力 A.可靠性增长试验所施加的应力强度可略高于使用时的应力强度，以不引起新的故障机理为限。如温度循环一般可以将略高于产品高温温度、略低于产品低温温度作为温度循环的上、下限温度，温度变化率可取5℃/min或10℃/min。循环周期时间根据温度变化率而定。 B.可靠性鉴定试验将产品工作高温温度作为试验温度。 C.ORT 试验按产品标准的工作高、低温温度进行各种功能和指标的检验。按产品标准的储运高、低温温度进行储运试验。 (二)试验条件③潮湿应力 A.可靠性增长试验预计受试产品在现场使用环境中会有明显的冷凝和结霜时，则在高温下应施加潮湿应力。B.可靠性鉴定试验同可靠性增长试验 C.ORT 试验应进行恒定湿热试验，湿度一般为90%~95%，高温温度一般为产品标准的工作高温温度。

软件可靠性与安全性分析、评估方法及建议

软件可靠性与安全性分析、评估方法及建议一、背景介绍随着产品技术的发展及数字化技术的应用，软件在产品中所占的比重越来越大，其规模和复杂性急剧增加，对产品的可靠性、安全性工作提出了严峻的考验。为保证软件可靠性，需要对软件进行可靠性测试和评估工作，从而尽早发现并改进软件中影响产品质量的缺陷，有效提高软件可靠性。为保障软件安全性，需要对软件进行安全性分析与验证工作。目前，随着GJB Z 161-2012 军用软件可靠性评估指南、GJB 900A-2012 装备安全性工作通用要求、GJB 102A-2012军用软件安全性设计指南、ARP4761与民用机载系统安全性评估流程及DO-178B/C机载系统合格审定过程中的软件考虑等标准的颁布实施，以及空军航定〔2012〕4号《航空军用软件定型测评进入条件评估准则》中明确提出关键软件在进入定型测评前必须具备《软件失效风险分析报告》；空军装型〔2010〕131号《空军重点型号软件工程化要求》中也明确提出在软件研制阶段中，必须要开展软件安全性分析与验证工作等规定。美国在70年代研制F/A-18飞机期间首次引入软件安全性技术。在研制F-22和F-35飞机时，则明确要求按照MIL-STD-882和DO-178B开展机载软件安全性工作。在民机领域，波音和空客均严格按照ARP-4761及DO-178B/C标准开展了软件安全性分析与验证，并作为适航审定的核心要素。在高铁、核工业、汽车、医疗等领域，同样要求按照IEC 61508、EN50128、IEC60880、IEC 61513、ISO 14971等标准，对构建高安全性软件做出严格规定。从上述可以看出，当前世界各国对于软件产品的可靠性评估、安全性分析验

可靠性增长试验

可靠性增长试验 1 概述 1.1基本概念众所周知，装备的可靠性是由设计决定的。但是，由于受到各种原因的影响，设计缺陷总是难免的，产品在研制阶段往往达不到用户的可靠性要求，因此必须开展可靠性增长活动。必须指出，可靠性增长活动不是针对设计低劣的产品的，而是针对经过认真设计仍然由于某些技术原因达不到要求的产品，而且可靠性增长活动比可靠性设计活动所需的资源和时间都多，因此，管理者往往只对通过可靠性设计评审的产品才安排可靠性增长计划。那种把可靠性水平寄托在增长活动上的态度是错误的。可靠性增长的核心是消除影响产品可靠性水平的设计缺陷。可靠性增长的关键是发现影响产品可靠性水平的设计缺陷。为此，必须通过试验或运行的途径来实现产品故障机理的检测。常见的可靠性增长有，一般性的可靠性增长和可靠性增长管理。一般性的可靠性增长，是指事前未给出明确的可靠性增长目标，对产品在试验或运行中发生的故障，根据可用于可靠性增长资源的多少，选择其中的一部分或全部实施纠正措施，以使产品可靠性得到确实提高的过程；它通常不制定计划增长曲线，也不跟踪增长过程，而是采用一两次集中纠正故障的方式，使产品可靠性得到提高。由于增长过程通常不能满足增长模型的限度条件，增长后的产品可靠性水平需要通过可靠性验证试验才能进行定量评估。可靠性增长管理，是指有计划有目标的可靠性增长工作项目，并非可靠性增长过程中的管理工作。它是产品寿命期内的一项全局性的、为达到预期的可靠性指标、对时间和资源进行系统安排、在估计值和计划值比较的基础上依靠新分配资源、对实际增长率进行控制的可靠性增长项目。可靠性增长管理有两个特点： a)有一个逐步提高的可靠性增长目标: 可靠性增长管理主要针对大型军事装备，把可靠性增长工作从工程研制阶段延伸到生产阶段或使用阶段，在阶段的转接处和阶段内部划分的小阶段的进出口处设定可靠性增长目标，形成逐步提高的系列目标。这就促使有关部门实施严格管理和为降低风险提供手段。b) 充分利用产品寿命期内的各项试验和运行记录: 除了可靠性试验之外，在产品寿命期内还有其它各种试验以及运行过程都可能产生故障信息，可以用于可靠性增长的故障机理检测，经过风险权衡后把其中的一部分纳入可靠性增长管理的范围，形成可靠性增长的整体，使产品可靠性逐步增长到预期目标。可靠性增长活动是一个连续完整的闭环控制过程。在此环中，首要任务是发现产品的设计缺陷——这主要是从试验、使用中发生的故障中发现；然后是对故障进行分析——重点研究重复性故障和关键故障发生的原因，当认定为设计缺陷后提出纠正这些设计缺陷的措施；接着是实施纠正措施——将修改设计的措施在少数产品(试验样品)上实施，并通过试验验证纠正措施的有效性；最后是修改技术文件和把纠正措施推广到同型号产品中去——这是落实可靠性增长活动的重要工作，是发挥可靠性增长试验效益的关键步骤。可靠性增长活动可以在工程研制阶段、生产阶段进行，甚至在使用阶段进行。按照有关标准的规定只在装备研制阶段才进行可靠性增长试验和增长工作，但从我国的实际情况出发，有不少已经装备部队多年的产品仍然对其进行可靠性增长试验和“可靠性补课工作”，并取得了显著成绩。这就是说，要根据产品的技术状况和可靠性水平去决定何时以何种形式开展可靠性增长活动。可靠性增长试验是可靠性增长活动的主要内容，是产品工程研制阶段单独安排的可靠性工作项目，成为工程研制阶段的组成部分。可靠性增长试验通常安排在工程研制基本完成之后和可靠性鉴定试验之前进行。此时，产品的性能与功能已经基本达到设计要求，产品结构与布局已经接近批生产的要求，故障信息的确实性已经较高，且此时故障纠正措施的实施所需资源和时间较少。使用阶段的可靠性增长活动可以利用产品的现场故障信息和现场使用状况记录来取代可靠性增长试验工作。 1.2可靠性增长试验的目的可靠性增长试验的目的是，在装备研制或生产阶段通过试验获得设计缺陷的信息，以便对其

几种常见软件可靠性测试方法综述及应用对比(精)

几种常见软件可靠性测试方法综述及应用对比上海交通大学陈晓芳 [摘要]软件可靠性测试是软件可靠性工程的一项重要工作内容,是满足软件可靠性要求、评价软件可靠性水平及验证软件产品是否达到可靠性要求的重要途径。本文探讨、研究了软件可靠性测试的基本概念,分析、对比了几种软件可靠性测试主要方法的优缺点。 [关键词]软件可靠性软件可靠性测试软件测试方法引言软件可靠性工程是指为了满足软件的可靠性要求而进行的一系列设计、分析、测试等工作。其中确定软件可靠性要求是软件可靠性工程中要解决的首要问题,软件可靠性测试是在软件生存周期的系统测试阶段提高软件可靠性水平的有效途径。各种测试方法、测试技术都能发现导致软件失效的软件中残存的缺陷,排除这些缺陷后,一般来讲一定会实现软件可靠性的增长,但是排除这些缺陷对可靠性的提高的作用却是不一样的。其中,软件可靠性测试能最有效地发现对可靠性影响大的缺陷,因此可以有效地提高软件的可靠性水平。软件可靠性测试也是评估软件可靠性水平,验证软件产品是否达到软件可靠性要求的重要且有效的途径。一、软件可靠性测试概念 “测试”一般是指“为了发现程序中的错误而执行程序的过程”。但是在不同的开发阶段、对于不同的人员,测试的意义、目的及其采用的方法是有差别的。在软件开发的测试阶段,测试的主要目的是开发人员通过运行程序来发现程序中存在的缺陷、错误。而在产品交付、验收阶段,测试主要用来验证软件产品是否达到用户的要求。或者说,对于开发人员,测试是发现缺陷的一种途径、手段,而对于用户,测试则是验收产品的一种手段。

二、软件测试方法软件测试方法有以下几个主要概念:白盒测试、黑盒测试、灰盒测试。白盒测试(W h ite-box testing或glass-box testing是通过程序的源代码进行测试而不使用用户界面。这种类型的测试需要从代码句法发现内部代码在算法,溢出,路径,条件等等中的缺点或者错误,进而加以修正。黑盒测试(B lack-box testing是通过使用整个软件或某种软件功能来严格地测试,而并没有通过检查程序的源代码或者很清楚地了解该软件或某种软件功能的源代码程序具体是怎样设计的。测试人员通过输入他们的数据然后看输出的结果从而了解软件怎样工作。通常测试人员在进行测试时不仅使用肯定出正确结果的输入数据,而且还会使用有挑战性的输入数据以及可能结果会出错的输入数据以便了解软件怎样处理各种类型的数据。灰盒测试(Gray-box testing就像黑盒测试一样是通过用户界面测试,但是测试人员已经有所了解该软件或某种软件功能的源代码程序具体是怎样设计的,甚至于还读过部分源代码,因此测试人员可以有的放矢地进行某种确定的条件或功能的测试。这样做的意义在于:如果你知道产品内部的设计和透过用户界面对产品有深入了解,你就能够更有效和深入地从用户界面来测试它的各项性能。 1、白盒测试白盒测试又称结构测试,透明盒测试、逻辑驱动测试或基于代码的测试。白盒测试是一种测试用例设计方法,盒子指的是被测试的软件,白盒指的是盒子是可视的,你清楚盒子内部的东西以及里面是如何运作的。白盒的测试用例需要做到: (1保证一个模块中的所有独立路径至少被使用一次; (2对所有逻辑值均需测试true和false;

串并联可靠性模型的应用及举例

上海电力学院选修课大型作业课程名称：机电系统可靠性与安全性设计报告名称：串并联可靠性模型的应用及举例院系：能源与机械工程学院专业年级：动力机械140101 学生姓名：潘广德学号：14101055 任课教师：张建平教授 2015年4月28日

浅谈串并联可靠性模型的应用并举例摘要详细阐述了机械可靠性工程中串并联可靠性模型的应用，并详细的举例说明。系统可靠性与组成单元的数量、单元可靠性以及单元之间的相互联接关系有关。以便于可靠性检测，首先讨论了各单元在系统中的相互关系。在可靠性工程中，常用可靠性系统逻辑图表示系统各单元之间的功能可靠性关系。在可靠性预测中串并联的应用及其广泛。必须指出，这里所说的组件相互关系主要是指功能关系，而不是组件之间的结构装配关系。关键词：机械可靠性串联并联混联应用举例 0前言学技术的发展，产品质量的含义也在不断的扩充。以前产品的质量主要是指产品的性能，即产品出厂时的性能质量，而现在产品的质量已不仅仅局限于产品的性能这一指标。目前，产品质量的定义是：满足使用要求所具备的特性，即适用性。这表明产品的质量首先是指产品的某种特性，这种特性反应这用户的某种需求。概括起来，产品质量特性包括：性能、可靠性、经济性和安全性四个方面。性能是产品的技术指标，是出厂时产品应具有的质量属性，显然能出厂的产品就赢具备性能指标；可靠性是产品出厂后所表现出来的一种质量特性，是产品性能的延伸和扩展；经济性是在确定的性能和可靠性水平下的总成本，包括购置成本和使用成本两部分；安全性则是产品在流通和使用过程中保证安全的程度。在上述产品特性所包含的四个方面中，可靠性占主导地位。性能差，产品实际上是废品；性能好，也并不能保证产品可靠性水平高。反之，可靠性水平高的产品在使用中不但能保证其性能实现，而且故障发生的次数少，维修费用及因故障造成的损失也少，安全性也随之提高。由此可见，产品的可靠性是产品质量的核心，是生产厂家和广大用户所努力追求的目标。 1串联系统可靠性模型的工作原理如果一个系统中的单元中只要有一个失效该系统就失效，则这种系统成为串联系统。或者说，只有当所有单元都正常工作时，系统才能正常工作的系统称为串联系统。设系统正常工作时间（寿命）这一随机变量为t，则在串联系统中，要使系统能正常工作运行，就必须要求每一个单元都能正常工作，且要求每一单元的正常工作时间都大于系统正常工作时间t。假设各个单元的失效时间是相互独立的，按照概率的乘法定理和可靠性定

可靠性增长与可靠性增长试验

众所周知，产品的可靠性是由设计决定的。但是，由于受到各种原因的影响，设计缺陷总是难免的，产品在研制阶段往往达不到用户的可靠性要求，因此必须开展可靠性增长活动。必须指出，可靠性增长活动不是针对设计低劣的产品的，而是针对经过认真设计仍然由于某些技术原因达不到要求的产品，而且可靠性增长活动比可靠性设计活动所需的资源和时间都多。 1、概述可靠性增长可从多个不同的角度来看，早期有关可靠性增长的一些工作主要集中在管理方面。1970年Selby和Miller研制的可靠性计划与管理（RPM）模型是联系可靠性要求和实施计划的管理工具，可帮助确定所需样品数和设计方案通过增长过程的成熟时间，并可监测进展情况，评价对原计划进行调整的必要性。但大多数情况下提及可靠性增长这一话题时，讨论的重点都是可靠性增长试验。一般而言，为了证明设计的正确性以及设计中使用的模型和分析工具的有效性，试验是开发的标准、必要部分。对于可靠性增长试验，大量的工作被用于研制各种统计模型，以便计划和跟踪通过试验所取得的可靠性增长。由于试验费用很高，因此自然会把很多精力放在研制好的模型和注重可靠性增长过程上。我们知道最常用的模型是Duane模型。Duane的观点是把整个重点放在试验中发现失效，然后通过重新设计予以排除。在笔者参加的某次“可靠性与风险分析先进课题”系列专题会议会议上，分组讨论中有一组的主题是“可靠性增长的范围和目的”。会上讨论了把试验作为实现可靠性增长首选方法的状况。其中一位成员提出，象卫星这样的产品，由于成本高，供试验的物品有限，因而极少可能进行那种和可靠性增长有关的试验。对这种系统如何实现可靠性增长呢？ 2、可靠性增长更广泛的

软件可靠性模型综述(完整资料).doc

【最新整理，下载后即可编辑】软件可靠性模型综述可靠性是衡量所有软件系统最重要的特征之一。不可靠的软件会让用户付出更多的时间和金钱, 也会使开发人员名誉扫地。IEEE 把软件可靠性定义为在规定条件下, 在规定时间内, 软件不发生失效的概率。该概率是软件输入和系统输出的函数, 也是软件中存在故障的函数, 输入将确定是否会遇到所存在的故障。软件可靠性模型，对于软件可靠性的评估起着核心作用，从而对软件质量的保证有着重要的意义。一般说来，一个好的软件可靠性模型可以增加关于开发项目的效率，并对了解软件开发过程提供了一个共同的工作基础，同时也增加了管理的透明度。因此，对于如今发展迅速的软件产业，在开发项目中应用一个好的软件可靠性模型作出必要的预测，花费极少的项目资源产生好的效益，对于企业的发展有一定的意义。 1软件失效过程 1.1软件失效的定义及机理当软件发生失效时，说明该软件不可靠，发生的失效数越多，发生失效的时间间隔越短，则该软件越不可靠。软件失效的机理如下图所示：

1）软件错误（Software error）：指在开发人员在软件开发过程中出现的失误，疏忽和错误，包括启动错、输入范围错、算法错和边界错等。 2）软件缺陷（Software defect）：指代码中存在能引起软件故障的编码，软件缺陷是静态存在的，只要不修改程序就一直留在程序当中。如不正确的功能需求，遗漏的性能需求等。 3）软件故障（Software fault）：指软件在运行期间发生的一种不可接受的内部状态，是软件缺陷被激活后的动态表现形式。 4）软件失效（Software failure）：指程序的运行偏离了需求，软件执行遇到软件中缺陷可能导致软件的失效。如死机、错误的输出结果、没有在规定的时间内响应等。从软件可靠性的定义可以知道，软件可靠性是用概率度量的，那么软件失效的发生是一个随机的过程。在使用一个程序时，在其他条件保持一致的前提下，有时候相同的输入数据会得到不同的输出结果。因此，在实际运行软件时，何时遇到程序中的缺陷导致软件失效呈现出随机性和不稳定性。所有的软件失效都是由于软件中的故障引起的，而软件故障是一种人为的错误，是软件缺陷在不断的测试和使用后才表现出来的，如果这些故障不能得到及时有效的处理，便不可避免的会

可靠性增长试验

可靠性增长试验 1 概述基本概念众所周知，装备的可靠性是由设计决定的。但是，由于受到各种原因的影响，设计缺陷总是难免的，产品在研制阶段往往达不到用户的可靠性要求，因此必须开展可靠性增长活动。必须指出，可靠性增长活动不是针对设计低劣的产品的，而是针对经过认真设计仍然由于某些技术原因达不到要求的产品，而且可靠性增长活动比可靠性设计活动所需的资源和时间都多，因此，管理者往往只对通过可靠性设计评审的产品才安排可靠性增长计划。那种把可靠性水平寄托在增长活动上的态度是错误的。可靠性增长的核心是消除影响产品可靠性水平的设计缺陷。可靠性增长的关键是发现影响产品可靠性水平的设计缺陷。为此，必须通过试验或运行的途径来实现产品故障机理的检测。常见的可靠性增长有，一般性的可靠性增长和可靠性增长管理。一般性的可靠性增长，是指事前未给出明确的可靠性增长目标，对产品在试验或运行中发生的故障，根据可用于可靠性增长资源的多少，选择其中的一部分或全部实施纠正措施，以使产品可靠性得到确实提高的过程；它通常不制定计划增长曲线，也不跟踪增长过程，而是采用一两次集中纠正故障的方式，使产品可靠性得到提高。由于增长过程通常不能满足增长模型的限度条件，增长后的产品可靠性水平需要通过可靠性验证试验才能进行定量评估。可靠性增长管理，是指有计划有目标的可靠性增长工作项目，并非可靠性增长过程中的管理工作。它是产品寿命期内的一项全局性的、为达到预期的可靠性指标、对时间和资源进行系统安排、在估计值和计划值比较的基础上依靠新分配资源、对实际增长率进行控制的可靠性增长项目。可靠性增长管理有两个特点： a) 有一个逐步提高的可靠性增长目标: 可靠性增长管理主要针对大型军事装备，把可靠性增长工作从工程研制阶段延伸到生产阶段或使用阶段，在阶段的转接处和阶段内部划分的小阶段的进出口处设定可靠性增长目标，形成逐步提高的系列目标。这就促使有关部门实施严格管理和为降低风险提供手段。 b) 充分利用产品寿命期内的各项试验和运行记录: 除了可靠性试验之外，在产品寿命期内还有其它各种试验以及运行过程都可能产生故障信息，可以用于可靠性增长的故障机理检测，经过风险权衡后把其中的一部分纳入可靠性增长管理的范围，形成可靠性增长的整体，使产品可靠性逐步增长到预期目标。可靠性增长活动是一个连续完整的闭环控制过程。在此环中，首要任务是发现产品的设计缺陷——这主要是从试验、使用中发生的故障中发现；然后是对故障进行分析——重点研究重复性故障和关键故障发生的原因，当认定为设计缺陷后提出纠正这些设计缺陷的措施；接着是实施纠正措施——将修改设计的措施在少数产品（试验样品）上实施，并通过试验验证纠正措施的有效性；最后是修改技术文件和把纠正措施推广到同型号产品中去——这是落实可靠性增长活动的重要工作，是发挥可靠性增长试验效益的关键步骤。可靠性增长活动可以在工程研制阶段、生产阶段进行，甚至在使用阶段进行。按照有关标准的规定只在装备研制阶段才进行可靠性增长试验和增长工作，但从我国的实际情况出发，有不少已经装备部队多年的产品仍然对其进行可靠性增长试验和“可靠性补课工作”，并取得了显著成绩。这就是说，要根据产品的技术状况和可靠性水平去决定何时以何种形式开展可靠性增长活动。可靠性增长试验是可靠性增长活动的主要内容，是产品工程研制阶段单独安排的可靠性工作项目，成为工程研制阶段的组成部分。可靠性增长试验通常安排在工程研制基本完成之后和可靠性鉴定试验之前进行。此时，产品的性能与功能已经基本达到设计要求，产品结构与布局已经接近批生产的要求，故障信息的确实性已经较高，且此时故障纠正措施的实施所需资源和时间较少。使用阶段的可靠性增长活动可以利用产品的现场故障信息和现场使用状况记录来取代可靠性增长试验工作。可靠性增长试验的目的

软件可靠性模型综述

软件可靠性模型综述可靠性是衡量所有软件系统最重要的特征之一。不可靠的软件会让用户付出更多的时间和金钱, 也会使开发人员名誉扫地。IEEE 把软件可靠性定义为在规定条件下, 在规定时间, 软件不发生失效的概率。该概率是软件输入和系统输出的函数, 也是软件中存在故障的函数, 输入将确定是否会遇到所存在的故障。软件可靠性模型，对于软件可靠性的评估起着核心作用，从而对软件质量的保证有着重要的意义。一般说来，一个好的软件可靠性模型可以增加关于开发项目的效率，并对了解软件开发过程提供了一个共同的工作基础，同时也增加了管理的透明度。因此，对于如今发展迅速的软件产业，在开发项目中应用一个好的软件可靠性模型作出必要的预测，花费极少的项目资源产生好的效益，对于企业的发展有一定的意义。 1软件失效过程 1.1软件失效的定义及机理当软件发生失效时，说明该软件不可靠，发生的失效数越多，发生失效的时间间隔越短，则该软件越不可靠。软件失效的机理如下图所示： 1）软件错误（Software error）：指在开发人员在软件开发过程中出现的失误，疏忽和错误，包括启动错、输入围错、算法错和边界错等。 2）软件缺陷（Software defect）：指代码中存在能引起软件故障的编码，软件缺陷是静态存在的，只要不修改程序就一直留在程序当中。如不正确的功能需求，遗漏的性能需求等。3）软件故障（Software fault）：指软件在运行期间发生的一种不可接受的部状态，是软件缺陷被激活后的动态表现形式。 4）软件失效（Software failure）：指程序的运行偏离了需求，软件执行遇到软件中缺陷可能导致软件的失效。如死机、错误的输出结果、没有在规定的时间响应等。

软件可靠性测试及其实践

软件可靠性测试及其实践 Software Reliability Te sting and Practice 北京航空航天大学工程系统工程系(100083)　陆民燕　陈雪松【摘要】软件可靠性测试是软件可靠性工程的一项重要工作内容,是满足软件可靠性要求、评价软件可靠性水平及验证软件产品是否达到可靠性要求的重要途径。本文探讨、研究了软件可靠性测试的基本概念,软件可靠性测试过程以及软件可靠性测试中的主要问题,还介绍了一个实际软件的可靠性测试工作。关键词:软件可靠性,软件可靠性测试,软件测试,软件运行剖面 Abstract :S oftware reliability testing is an important task in s oftware reliability engineering.It serves as the main means to achieve the s oftware reliability requirements ,evaluate s oftware reliability levels and dem onstrate whether a s oftware product has achieved its reliability requirement.This paper investigates the basic concepts of s oftware reliability testing and s oftware reliability testing procedures ,discusses the key issues in it.A practical w ork on a real s oftware product is als o presented.K ey w ords :softw are reliability ,softw are reliability test 2 ing ,softw are testing ,operational profile 软件可靠性工程是指为了满足软件的可靠性要求而进行的一系列设计、分析、测试等工作。其中确定软件可靠性要求是软件可靠性工程中要解决的首要问题。软件可靠性要求可以包括定性定及量要求。软件可靠性测试是在软件生存周期的系统测试阶段提高软件可靠性水平的有效途径。各种测试方法、测试技术都能发现导致软件失效的软件中残存的缺陷,排除这些缺陷后,一般来讲一定会实现软件可靠性的增长,但是排除这些缺陷对可靠性的提高的作用却是不一样的。其中,软件可靠性测试能最有效地发现对可靠性影响大的缺陷,因此可以有效地提高软件的可靠性水平。软件可靠性测试也是评估软件可靠性水平,验证软件产品是否达到软件可靠性要求的重要且有效的途径。 1　软件可靠性测试概念 “测试”一般是指“为了发现程序中的错误而执行程序的过程”。但是在不同的开发阶段、对于不同的人员, 测试的意义、目的及其采用的方法是有差别的。在软件开发的测试阶段,测试的主要目的是开发人员通过运行程序来发现程序中存在的缺陷、错误。而在产品交付、验收阶段,测试主要用来验证软件产品是否达到用户的要求。或者说,对于开发人员,测试是发现缺陷的一种途径、手段,而对于用户,测试则是验收产品的一种手段。根据测试用例选取原则的不同,测试可分为黑盒测试方法和白盒测试方法两大类。黑盒测试方法是指按照软件需求生成测试用例对软件进行测试的方法,黑盒测试不关心程序是如何实现的;而白盒测试方法则是指根据程序的结构生成测试用例对软件进行测试的方法。软件可靠性测试是指为了保证和验证软件的可靠性要求而对软件进行的测试。其采用的是按照软件运行剖面(对软件实际使用情况的统计规律的描述)对软件进行随机测试的测试方法。通过软件可靠性测试可以达到以下目的: (1)有效地发现程序中影响软件可靠性的缺陷,从而实现可靠性增长:软件可靠性是指[4] “在规定的时间内,规定的条件下,软件不引起系统失效的能力,其概率度量称为软件可靠度。”软件的“规定的条件”主要包括相对不变的条件和相对变化的条件,相对不变的条件如计算机及其操作系统;相对变化的条件是指输入的分布,用软件的运行剖面来描述。按照软件的运行剖面对软件进行测试一般先暴露在使用中发生概率高的缺陷,然后是发生概率低的缺陷。而高发生概率的缺陷是影响产品可靠性的主要缺陷,通过排除这些缺陷可以有效地实现软件可靠性的增长。 (2)验证软件可靠性满足一定的要求:通过对软件可靠性测试中观测到的失效情况进行分析,可以验证软件可靠性的定量要求是否得到满足。 ③估计、预计软件可靠性水平:通过对软件可靠性测试中观测到的失效数据进行分析,可以评估当前软件可靠性的水平,预测未来可能达到的水平,从而为开发管理提供决策依据。软件可靠性测试中暴露的缺陷既可以是影响功能需求的缺陷也可以是影响性能需求的缺陷。软件可靠性测试方法从概念上讲是一种黑盒测试方法,因为它是面向需求、面向使用的测试,它不需要了解程序的结构以及如何实现等问题。 ?84? 《测控技术》2000年19卷第5期

软件可靠性工程范文

软件可靠性工程 1.软件可靠性定义 1.1.广义是指一切旨在避免、减少、处理、度量软件故障（错误、缺陷、失效）的分析、设计、测试等方法、技术和实践活动。于是有诸多相关术语，如软件可靠性度量、软件可靠性设计、软件可靠性建模、软件可靠性测试、软件可靠性管理等。 1.2.狭义指软件无失效运行的定量度量，尤其是那些面向用户的定量度量。主要有： ?软件可靠度：表示软件在规定的运行环境中和规定的运行时间内无失效运行的机会。软件无失效运行的机会多以概率度量，但也可以模糊数学中的可能性加以度量，有时也在数据域上将软件可靠度表示为软件成功执行一个回合的概率。 ?软件失效强度：其物理解释是单位时间内软件发生失效的机会。在概率范畴内，它与软件可靠度有明确的数学关系（R(t)＝1－F(t)，R(t)为可靠度，F(t)为失效强度）。 ?软件平均失效时间（MTTF）：表示软件投入运行到出现一个新失效的时间。上述度量与硬件可靠性中的相应概念本质上是一致的。 “失效”是指程序的功能在某方面没有达到用户的需求。“没有像用户需求的那样工作”是一个很广的定义。因此，可靠性结合了与程序执行相关联的所有属性。例如，它包括正确性、安全性和可使用性的操作方面，以及对用户的友好性。请注意，安全性实际上是软件可靠性的一个特殊子类。可靠性不包括可移植性、可修改性或文档的可理解性。

可靠性是面向用户的而不是面向开发人员的。可靠性与操作有关，而不是与程序的设计有关，因此可靠性是动态的，而不是静态的。可靠性考虑问题出现的频率，直接与操作经验和在经验中错误的影响相关。因此，可以很容易地将可靠性与成本联系起来。可靠性很适合检查发展趋势的重要性、设定目标和预测什么时候可以达到目标。可靠性使人们可以使用同样的术语对硬件和软件的系统可靠性进行分析，而在真实系统中硬件和软件都同时存在。所以，可靠性度量比错误度量要有用得多。 2.软件可靠性工程的研究范围软件可靠性工程涉及以下四方面活动和有关技术： 2.1.软件可靠性分析进行软件可靠性的需求分析、指标分配、故障树分析、失效模式和影响分析、软件开发过程中有关软件可靠性的的特性分析、……等。 2.2.软件可靠性设计和实现进行防错设计、容错设计、检错设计、纠错设计、故障恢复设计、软件可靠性增长、……等。 2.3.软件可靠性测量、测试和评估在软件生存周期各阶段进行有关软件可靠性设计、制造和管理方面的属性测量，进行基于软件运行剖面的测试用例随机输入的软件测试、软件可靠性预计、软件可靠性估计、软件可靠性验证、……等。 2.4.软件可靠性管理确定影响软件可靠性的因素，制定必要的设计和实现准则以及对软件开发各阶段软件可

可靠性试验设计与分析5

第四章(45) 可靠性试验设计与分析 §4．4可靠性增长试验（Reliability Growth Test）一、概述可靠性增长：通过改正产品设计和制造中的缺陷，不断提高产品可靠性的过程。产品试制阶段，由于设计缺陷与工艺上的不成熟，其可靠性一定会远低于预计的标准，通过试验发现故障，通过机理分析找出故障源，通过再设计与工艺的更改，以达到消除故障的目的，保证研制期间的可靠性达到预期的指标。可靠性增长是不断反复设计、试验、故障、纠正这样一个循环过程。是为达到可靠性增长目的而执行可靠性秩序中所采用的一种试验方法。可靠性增长的三个主要因素： 1).通过分析和试验找出产品的潜在故障源。 2).将存在问题（返馈），采取纠正措施更改设计。 3).对改进后的产品重新进行试验。图4.23 可靠性增长过程二、可靠性增长试验目的：通过试验诱导出设计不良或工艺不成熟而引起的潜在故障，通过机理分析找出问题，在设计与工艺上加以纠正，从而达到可靠性增长目的。可靠性增长试验耗费的资源和时间比较多，试验总时间通常为预期的MTBF目标值的5~25倍，所以也并不是所有产品都适宜于安排可靠性增长试验。其试验大纲按照试验、分析、纠正(Test, Analysis And Fix test简称TAAF)这一过程来制定，为此要选定一个可靠性增长的模型，以便确定试验计划时所需考虑的因素。 1、可靠性增长模型目前在可修产品的增长试验中，普遍使用的杜安(Duane)模型。有时为了使杜安模型的

适合性和最终评估具有较坚实的统计学依据，可用AMSAA 模型作为补充。杜安模型是用于飞机发动机和液压机械装置等复杂可修产品的增长试验的。模型未涉及随机现象，是确定性模型，即工程模型，而不是数理统计模型。其基本假设：只要不断进行可靠性试验，系统可靠性增长（用MTBF 的提高表示）与累积试验时间在双对数纸上成线性关系，直线的斜率是可靠性增长率的一个度量。图4.24 可靠性增长曲线上述描绘了杜安可靠性增长模型。其增长率范围在0.3~0.7之间，若增长在0.3以下，说明纠正措施不够有力，在0.7以上表明采用了强有力的纠正措施。从曲线上还可表明，制定可靠性增长大纲所需要的四个因素： (1).系统固有的MTBF 值P q 与要求的MTBF 值s θ关系： p θ在设计时用预测的方法确定，而s q （可接受值）比P q 低些，这是验证试验之前应增长到的值。 (2).增长曲线的起始MTBF 值0q ：当P q 预期值为200h ￡时，增长线以100试验小时（横坐标）与10%P q （纵坐标）为起始点。当200P h q >，则以100h 试验与50%P q 为起始点。 (3).关于MTBF 增长率a ：取决于大纲要求，如制定合理并执行严格，增长率可达0.6，没有特殊考虑时可取0.1a =。 (4).增长所要求的总时间：增长线与指标要求的MTBF 值的水平线交点所对应的总试验时间即为预计总试验时间。美国军用标准有个试验指南：当固定的试验持续时间为规定的MTBF （s q ）的10～25倍时，该时间完全可以满足达到50～2000hMTBF 内预期的设备可靠性增长需要。当规定的MTBF 在2000h 以上时，其持续试验时间取决于设备的复杂性和大纲要求，但至少应是要求的MTBF 的一倍。无论任何情况下，持续时间试验都不得少于2000h 或不多于10000h 。

可靠性建模资料整理

软件可靠性建模 1模型概述 1.1软件可靠性的定义 1983年美国IEEE计算机学会对“软件可靠性”作出了明确定义，此后该定义被美国标准化研究所接受为国家标准，1989年我国也接受该定义为国家标准。该定义包括两方面的含义：（1）在规定的条件下，在规定的时间内，软件不引起系统失效的概率；（2）在规定的时间周期内，在所述条件下程序执行所要求的功能的能力；其中的概率是系统输入和系统使用的函数，也是软件中存在的故障的函数，系统输入将确定是否会遇到已存在的故障（如果故障存在的话）。软件失效的根本原因在于程序中存在着缺陷和错误，软件失效的产生与软件本身特性、人为因素、软件工程管理都密切相关。影响软件可靠性的主要因素有软件自身特性、人为因素、软件工程管理等，这些因素具体还可分为环境因素、软件是否严密、软件复杂程度、软件是否易于用户理解、软件测试、软件的排错与纠正以及软件可靠性工程技术研究水平与应用能力等诸多方面。 1.2软件可靠性建模思想建立软件可靠性模型旨在根据软件可靠性相关测试数据，运用统计方法得出软件可靠性的预测值或估计值，下图给出了软件可靠性建模的基本思想。

图软件可靠性建模基本思想从图中可以看出软件失效总体来说随着故障的检出和排除而逐渐降低，在任意给定的时间，能够观测到软件失效的历史。软件可靠性建模的目标如下：（1）预测软件系统达到预期目标所还需要的资源开销及测试时间；（2）预测测试结束后系统的期望可靠性。1.3软件可靠性建模基本问题软件可靠性建模需要考虑以下基本问题：（1）模型建立模型建立指的是怎样去建立软件可靠性模型。一方面是考虑模型建立的角度，例如从时间域角度、数据域角度、将软件失效时刻作为建模对象，还可以将一定时间内软件故障数作为建模对象；另一方面是考虑运用的数学语言，例如概率语言。（2）模型比较在软件可靠性模型分类的基础上，对不同的模型分析比较，并对模型的有效性、适用性、简洁性等进行综合权衡，从而确定出模型的适用范围。（3）模型应用软件可靠性模型的应用需要从以下两方面考虑：一是给定了软件的开发计划，如何选择适当的模型；二是给定了软件可靠性模型，如何指导软件可靠性工程实践。软件系统的失效历史可以通过对测试得到的失效数据分析获得，而实际情况中，人们最为关注的是软件未来的失效趋势。软件可靠性模型基本都是建立在一定的假设基础之上，所以，即使花费了大量的时间和精力对软件的可靠性进行预计，也只是一种预测，这

可靠性增长技术的现状及发展

中国航空工业总公司可靠性和环境试验管理中心(北京100083) 摘要可靠性增长技术是随着可靠性增长定量化(增长模型)的研究及其在工程中的应用而发展起来的。本文通过大量的国内外资料，在分析了国内外增长模型研究的基础止，举例综述了国内外可靠性增长技术在工程中应用的现状和发展，以及目前结合我国国情开展的可靠性增长工程活动。关键词可靠性增长历史现状发展一、可靠性增长和增长试验产品固有可靠性是设计出来的，制造出来的.由于产品复杂性的不断提高及新技术的不断应用，产品更新换代加速.设计一种产品，需要有一个不断深化认识、逐步改进和完善的过程。初始产品(或称为样机)在试验或运行中故障较多，存在许多设计和工艺方面的缺陷。通过有计划地改进设计和工艺，根除故障产生的原因，提高产品的可靠性水平，逐步达到预期的目标。这种有计划的按照“试验一分析一改进”(Test, Analysis and Fix，缩写为TAAF)的程序，使产品可靠性获得提高的过程，称之为可靠性增长。可靠性增长可分两种情况:在研制过程中的增长主要是通过改进设计来达到的;在生产过程中主要是通过对产品的筛选或老炼过程排除产品中的不良元器件、部件和工艺缺陷而得到增长。当然，使用中又会进一步发现设计和制造工艺的不足之处，同时还会不断地收到使用中故障信息的反馈，通过改进产品使可靠性进一步提高。实现可靠性增长的关键在于发现故障，分析原因，并采取纠正措施。为了发现故障，就要对产品进行试验或运行，所以实现可靠性增长离不开可靠性增长试验。可靠性增长试验是在产品研制过程中有计划地实行试验、分析及解决问题的一个过程.在这个过程中，产品处在实际环境、模拟环境或加速变化的环境下经受试验，以暴露设计中的缺陷。所以，可靠性增长试验是在试验中激发产品故障、进行分析后采取有效的纠正措施、使产品固有可靠性得到不断提高的一种试验。由于可靠性增长试验是通过TAAF过程来提高产品的可靠性，所以它包含了对产品性能的监测、故障分析及其对减少故障再现的设计改进措施的检验。另外，可靠性增长试验本身只能暴露问题，不能提高产品固有可靠性，只有采取纠正措施，防止故障再现才能达到可靠性增长的目的。可见，可靠性增长试验有两个作用:一是通过试验发现故障;二是判断或验证所采取的纠正措施的有效性. 可靠性增长试验是可靠性工程的一个组成部分。从可靠性增长的工程原理可以看出，可靠性增长试验离不开其它的可靠性工作，而其它工作的结果又将体现于或被利用于可靠性增长试验之中，这样才能将这些工作转化为工程上的现实效益。可以说，可靠性增长试验是联结、带动各项可靠性工作的一条主线。它可以作为一种手段来全面审查所有的可靠性活动，因此它是可靠性工程中最重要的项目之一。应该指出，要注意区分可靠性增长试验和可靠性增长，可靠性增长试验是在产品研制阶段进行，但产品的可靠性增长贯穿了产品的设计、制造、使用和维修的全过程，即在研制阶段结束后，仍然存在可靠性增长的要求。通过使用和维修中发现问题，不断改进产品，提高产品的可靠性。当然，在可靠性增长及试验的概念提出来之前，就已经有人自觉或不自觉地按照TAAF的过程来提高产品的可靠性，尽管那时还没有出现“可靠性增长”一词.然而，这种做法可以说只停留在定性水平上，还没有达到定量要求.应该说，定量化的研究是从可靠性增长数学模型的研究开始的。总之，可靠性增长是产品可靠性水平不断提高的一种现象或状态。产生这种现象的根源是可靠性增长试验等一系列活动。这种活动造成的可靠性增长现象是有规律可循的，即可以用可靠性增长模型这种数学形式来描述。有了增长模型，研制活动的规划者和管理者就可以利用较为准确的定量方法来计算和评估研制对象的可靠性状态，并且在一定范围内预测末来的可靠性状态，以便就拥有的各项资源作出系统的安排.

可靠性试验

可靠性试验一、单项选择题(每题的备选项中，只有1个最符合题意) 1. 可靠性试验可分为工程试验和。 A.统计试验 B.可靠性鉴定试验 C.可靠性验收试验 D.可靠性增长试验答案：A [解答] 可靠性试验一般可分为工程试验和统计试验。工程试验包括环境应力筛选试验和可靠性增长试验；统计试验包括可靠性鉴定试验、可靠性测定试验和可靠性验收试验。 2. 可靠性试验一般可分为工程试验和统计试验。工程试验包括、可靠性增长试验。 A.环境应力筛选试验 B.可靠性鉴定试验 C.可靠性测定试验 D.可靠性验收试验答案：A 3. 在批生产阶段早期环境应力筛选应。 A.100%进行 B.抽样进行 C.视产品质量情况具体决定试验样本量 D.条件不具备可部分进行

答案：A [解答] 不论是产品开发阶段，还是批生产阶段早期，环境应力筛选在元器件、组件、部件等产品层次上都应100%的进行。在批生产阶段后期，对组件级以上的产品可根据其质量稳定情况抽样进行。 4. 下列各项活动中，可以提高产品的固有可靠性。 A.可靠性计算 B.换件维修 C.改进设计 D.可靠性试验答案：C 5. 剔除由不良元器件，零部件或工艺缺陷引起的早期故障的方法是。 A.环境应力筛选试验 B.常规的方法 C.目视检查 D.任何方法或试验答案：A [解答] 环境应力筛选试验是通过在产品上施加一定的环境应力，以剔除由不良元器件、零部件或工艺缺陷引起的产品早期故障的一种工序或方法。这种早期故障通常用常规的方法和目视检查等是无法发现的。 6. 下列关于可靠性增长试验的说法正确的是。 A.可靠性增长试验是一个在规定的环境应力下，为暴露产品薄弱环节而进行的试验 B.可靠性增长试验是通过发现故障、分析和纠正故障以及对纠正措施的有效性进行验证以提高产品可靠性水平的过程