当前位置：文档库 › Center for Pervasive Comm.

Center for Pervasive Comm.

An Acoustic Identi?cation Scheme for Location Systems

Raja Jurdak

UCI/School of Information

and Computer Science Center for Pervasive Comm.

and Computing

rjurdak@https://www.wendangku.net/doc/086936193.html,

Cristina Videira Lopes

UCI/School of Information

and Computer Science

Center for Pervasive Comm.

and Computing

lopes@https://www.wendangku.net/doc/086936193.html,

Pierre Baldi

UCI/School of Information

and Computer Science

California Institute for

Telecomm.and Information

Technology Cal-(IT)2

pfbaldi@https://www.wendangku.net/doc/086936193.html,

Abstract

Many pervasive computing applications require location awareness in order to successfully integrate technology into our daily lives.A location system consists of a group of sensors that determine the position of a mobile user and provide the user with useful context-speci?c information. Locating the user requires signal exchanges between the user’s mobile device and the sensors.This paper considers an acoustic-based location system for pervasive computing applications.The system is comprised of a set of micro-phones connected to a central server.Mobile users produce acoustic signals through standard speakers,which are al-ready available in most mobile devices,to perform location operations through the system.We focus on the design of robust acoustic signals using multi-frequency symbols that serve both in locating and uniquely identifying the user.We conduct experiments at distances between1and17ft to ex-plore the ability of the server to recognize and decode sig-nals originated by different users in the same general area.

1.Introduction

Most pervasive computing applications require location awareness in order to successfully integrate technology into our daily lives.A location system consists of a set of sensors that determine the position of a mobile user and provide the user with useful context-speci?c information.Locating the user requires signal exchanges between the user’s mobile device and the sensors.

Here,we present work that is part of a project to de-sign an acoustic location system that combines the bene?ts 00–7803–8577-2/04/$20.00c 2004IEEE of a computer network with the intrinsic worth of location-awareness for roaming users in the physical world.In this context,a user equipped with a roaming device will be able to obtain on-demand user and location-speci?c information. Applications for this system include cheap and easily de-ployable location systems,extensions to web protocols,and location-dependent multi-user games.Areas of deployment include places that have been to a large extent abstracted on the Internet,such as retail stores,shopping malls,mu-seums,amusement parks,and other places where people seek related information,and eventually pay for products and services.The main contribution of this paper is to ex-plore the design of robust acoustic signals for locating and identifying multiple users within the coverage area of the location system.

The rest of the paper is organized as follows.Section2 reviews the previous related work.Section3provides an overview of the operation of the acoustic location system. Section4discusses the frequency encoding and transmis-sion of symbols which are the building blocks of user ID’s in our system.Section5discusses mainly the mechanisms involved in detecting the frequency content and in decoding the symbols contained in the received acoustic signal.Sec-tion6presents the experiments we performed to validate this encoding scheme for acoustic location and identi?ca-tion.Section7discusses the results of the experiments and concludes the paper.

2.Related Work

Several location systems[1]have been proposed to pro-vide location awareness in a ubiquitous computing environ-ment.In Active Badge[2],which is one of the earliest pro-posed location systems,users wear badges that emit diffuse infrared signals.Pre-installed sensors detect the infrared signals and report them to a central server to determine the

user’s location.Infrared waves have several undesirable features for location systems,including interference from ?orescent lighting and sunlight.

Other location systems,such as Active Bat[3]and Cricket[4],rely on ultrasound signals.Active Bat’s archi-tecture is similar to Active Badge in that it requires mobile users to wear ultrasound tags,and ceiling-mounted ultra-sound receivers capture the tag’s signal and report it to the central server.Active Bat uses an ultrasound time-of-?ight lateration technique,in which the user sends both an ultra-sound and radio signal,and the system computes the differ-ence in arrival times between the two signals to determine the user’s position.Cricket enhances Active Bat by using the radio signal arrival time to narrow the time window in which arriving signals are considered.Dolphin[5]is an-other ultrasound positioning system that has a distributed ?avor.In Dolphin,the location of only a few nodes is known,and the remaining nodes can infer their own loca-tion based on the location of the reference nodes.Nodes in Dolphin also send messages periodically to advertise their position and to maintain synchronization.

Because of their reliance on technologies such as in-frared and ultrasound,most existing location systems often require the user to carry additional hardware such as badges or tags.Requiring additional specialized hardware on the user side introduces cost and feasibility issues,which in turn limit the large-scale proliferation of existing systems.

Other location systems proposed the use of hardware that is already found in mobile devices.RADAR[6]uses the signal to noise ratio and signal strength of a mobile user’s IEEE802.11[7]transmissions to locate the user in a2di-mensional environment.One drawback of the RADAR sys-tem is its assumption that the mobile device is equipped with IEEE802.11,which does not apply to all mobile de-vices.Security and privacy also arise as important issues in radio frequency location systems:the system can track the user without the user knowing it.

Unlike WLAN technologies that are not available in smaller mobile devices and that are protocol dependent,the acoustic interface is available in virtually all mobile de-vices and is universally compatible.Acoustic technology has been recently considered for ubiquitous computing and communications applications.Lopes and Aguiar[8]have explored the use of musical sounds or other familiar sounds for low bit rate communications using hardware that is read-ily available in desktop computers,palm devices,memo recorders,televisions and other electronic devices.Simi-larly,our aim is to use sounds that are easily reproducible by most mobile devices for indoor location.

The work in[9]considers an outdoor location system based on a network of acoustic sensors to provide high lo-cation accuracy at considerable monetary cost for military and scienti?c applications.The system in[9]assumes a fully distributed self-organizing architecture where the sen-sors discover the topology and integrate into the network, which adds complexity and cost to the sensors themselves. In contrast,the design goal of our system is the develop-ment of an indoor acoustic positioning system with reason-able accuracy for cheap and easy deployment.Our system adopts a centralized topology,where microphones are only input devices through which the acoustic signals are relayed to the centralized server.Furthermore,the system in[9] employs complex algorithms for sensor synchronization,as well as beamforming techniques to determine the direction from which the signal arrives at the microphone.On the other hand,our system uses simple UDP sockets for tempo-rary synchronization between the server and sensors,which eliminate the overhead for continuous synchronization be-tween the sensors.Our system also replaces complex beam-forming techniques with basic triangulation1at the server to reduce the latency of a location operation.

The work in[10]also proposes the use of acoustic waves for indoor location systems,as well as for low bit rate com-munications.In[10],Madhavapeddy et al.consider several audio modulation techniques,including Dual Tone Multi-Frequency(DTMF),melodic sounds,and inaudible signals at the border of the acoustic range.In Madhavapeddy’s acoustic location architecture,one of several listeners de-tect the acoustic signal and report the signal characteristics to a central server.Our system uses a similar architecture to determine the user’s location but with?ner granularity. While the location system in[10]aims to identify the room in which the user is located,our system employs acoustic signals to locate a user’s approximate position within the room.

3.System Overview

3.1.Signal Description

Mobile users in Madhavapeddy’s acoustic location scheme emit a single tone signal at800Hz.From a tech-nical viewpoint,we believe that the use of a single acous-tic tone to locate and identify users is insuf?cient because frequency-speci?c noise impulses(or their harmonics)in some indoor environments may prevent the listeners from capturing the correct signal from the user.Furthermore,us-ing a single tone signal does not support multiple simultane-ous users.Madhavapeddy’s main motivation in using a sin-gle frequency for the location operation is that the base ca-pability of mobile phones is to produce monophonic tones.

1Triangulation at the server uses multiple distance measurements from different microphones[1].Around each microphone,the server considers a sphere with the reported distance of that particular microphone.The location of the user is then computed as the intersection of the spheres around the microphones that report the shortest distances.

are becoming increasingly popular.Other mobile devices such as handheld devices and pocket PC’s are already ca-pable of producing polyphonic tones.Thus,our acoustic location and identi?cation scheme combines two variants of Madhavapeddy’s proposed modulation techniques:

1.We use3frequencies to represent each symbol in our

alphabet,while DTMF uses2frequencies to represent each symbol in touch/tone https://www.wendangku.net/doc/086936193.html,ing3frequen-cies provides redundancy that increases the signal’s ro-bustness to frequency-selective interference and fad-ing.

2.We choose the frequencies that belong to the same ma-

jor musical scale,as suggested in[8],to represent each symbol in our alphabet.

3.2.Architecture

The system architecture includes a set of listening de-vices(microphones)that are connected to a central server, as shown in Figure1.Figure1shows a simple scenario where four microphones,which are mounted in a grid for-mation on the ceiling of a room,are also connected to a server that stores a map of the room.The distance between microphones is assumed to be suf?ciently small to ensure that at least three of the microphones detect the acoustic signals from any area within the room.

3.3.Location Operation

We illustrate the sequence of events involved in a loca-tion operation through the following example.Consider a customer that is seeking directions to the appliances section

store,which already has a deployed

such as IEEE802.11[7]or Blue-

customer’s mobile device is also assumed

data communications capability.Conse-

data network can assign the mobile device

DHCP so that the location system can

on the data network.

can issue a location query through his

to request directions,and the server subse-

the requested directions.However,the

determine the current location(context)of

the store before providing the customer

to the appliances section.To en-

to locate the user,the mobile device sends

4)in the form of acoustic signals.At

ceiling-mounted microphones detect these

In the topology of Figure1,microphones

the user’s acoustic ID.In general,each

detects the signal reports it to the central

should then estimate the distance of the user from each https://www.wendangku.net/doc/086936193.html,ing absolute signal strength to estimate distance is not reliable in this application be-cause the amplitude of the acoustic signal varies among mo-bile devices.Instead,the server uses time measurements to estimate the distance of the user from each microphone. The steps involved in distance estimation are the following: 1.While the server is constantly listening for client re-

quests,the client on the mobile device noti?es the server through the wireless data network that it is about to send its acoustic ID to initiate a location query.

2.The client then immediately sends the acoustic signal

encoding of its ID.The server can compute the latency between the end of the noti?cation signal received on the data network and the beginning of the acoustic sig-nal.Based on this latency and the propagation char-acteristics of both signals,the server estimates the dis-tance of the user from a single microphone.

Our preliminary experiments have shown that the server can determine the distance of the sound source located at a distance of up to22ft with an error of6inches for a sin-gle frequency signal.After estimating the distance of the user from each microphone,the server uses triangulation to determine the current location of the user,as shown in Fig-ure1.Once triangulation yields an estimate of the user’s location,the server uses its locally available store map to determine the proper directions from the user’s current lo-cation to the appliances section.The server can then send directions to the user through the wireless data network.

In other settings where the mobile device does not have a separate communications capability,the system could use relative received signal strengthes from various micro-phones to estimate the user’s location.Subsequently,the

system can use a modulated acoustic signal,as described in[8]and[10],to provide directions to mobile devices that are equipped with microphones.

4.Signal Encoding and Transmission

4.1.Acoustic ID’s

We now focus on the structure of the acoustic signal. The signal encoding serves both in locating the user and in uniquely identifying the user among all current users within the room.The purpose of embedding a unique ID into the location signal is twofold.First,in the absence of a sepa-rate data network with its own addressing scheme,acoustic ID’s provide a way to uniquely identify users in the cover-age area and to route information to users.Secondly,if two or more users simultaneously issue location queries,ID’s enable the server to determine the source of each query.

Assigning unique ID’s requires mobile devices to regis-ter with the server upon entering the coverage area of the location system.During the registration process,the user requests an acoustic ID along with an IP address to iden-tify him/her uniquely among other users within the cover-age area.Once the system assigns the user an ID and an IP address,the user can roam within the coverage area and use the assigned ID to initiate location operations.

4.2.ID encoding

We use a coding scheme similar to the DTMF scheme to encode each symbol.DTMF encodes each digit using2 separate frequencies in the range697to1633Hz.Frequen-cies in DTMF are classi?ed as high and low frequencies, and DTMF encodes each digit using one low and one high frequency.As a result,some pairs of digits in DTMF have one frequency in common.Consequently,if a receiver de-tects only one of the two frequencies of a DTMF symbol, it is impossible to decode that symbol.This becomes more of an issue for wireless acoustic communication,which has higher signal losses than the intended application area of DTMF in wired phone lines.Even if no frequencies are lost,the common frequencies among digits in DTMF may prevent receiver from deterministically identifying2digits received simultaneously over the air.Thus,we propose en-hancements to DTMF in our coding scheme to provide a more robust signal that meets the needs of wireless acous-tics.

4.2.1Symbols and Frequencies

Our coding scheme uses an alphabet of3symbols,which is easily extendible to6or more symbols.These symbols

Digit Freq1Freq2Freq3Scale

1279435204186F Major

2313639514699G Major

3248929603729D#Major Table1.Frequency Encoding of Symbols

form the building block of user ID’s,since each combina-tion of symbols represents a unique user ID.Each symbol is encoded using3frequencies,and unlike DTMF,there are no common frequencies among any two symbols.The choice of using3frequencies to encode each symbol stems from the fact that the microphones may not detect all the transmitted frequencies.This redundancy,together with the property that each frequency correlates only to one symbol, ensures that symbols are correctly decoded by the server even if some frequency components are lost.Our scheme requires the receiver to detect two out of the three frequen-cies2that represent a symbol to decode that symbol.

We use frequencies in the range of2200to4700Hz, since this band is less susceptible to indoor background noise,which we observed to be at frequencies below2Khz. Another motivation for using frequencies in the range2.2 Khz-4.7Khz is that the speakers of many mobile devices operate well in that range.

Unlike other communication technologies,acoustic sig-nalling and communications can be perceived by humans. As a result,any wireless acoustic system should ensure that the emitted signals are pleasant,or at least tolerable,to the human auditory system.To address this issue,each set of three frequencies that represent a symbol in our system be-long to the same major musical scale,so that sending an audio signal for a single symbol does not annoy users.

As mentioned earlier,the coding scheme encodes3sym-bols using9frequencies,where each symbol is the sum of three sinusoidal signals at frequencies which lie on the same major scale.Table1indicates the frequencies that encode each symbol,and the musical scale on which each set of three frequencies fall.Note that all frequencies in Table1 are separated by at least150Hz,so that any frequency shifts due to hardware variations do not result in false ID’s.

In addition to the musical scales in Table1,we are ex-ploring several other musical scales,such as the blues major scale,the blues minor scale and the?amenco scale,for pro-ducing acoustically pleasant ID’s.

2The requirement of receiving two out of three frequencies for each digit may be overly redundant for some environments with predictable noise patterns.This requirement could be relaxed to one out of three fre-quencies,which would increase the range of detection of a user,but it would also increase the probability of error in location and identi?cation. The modi?ed scheme could also triple the number of users by using one frequency to identify each user,at the cost of decreased reliability.

Samples

A m p l i t u d e

Figure 2.Signal shape of the transmitted ID [1,2,3].

4.2.2Reverberation

Ideally,a mobile device would send each symbol immedi-ately after the preceding symbol.In reality however,acous-tic waves in a room undergo reverberation due to the re?ec-tion of sound within the room [12].Reverberant sound in a room dies away as the sound energy is absorbed by mul-tiple interactions with the surfaces of the room.Thus,the ID encoding scheme requires a guard time between each two symbols to reduce the effects of reverberation from one symbol to the next.Figure 2plots the signal of ID [1,2,3]with the symbols separated by 9.7millisecond guard times.Figure 3shows the same ID as it is captured by the receiver.The symbol boundaries in Figure 3are still visible at the guard times,where only reverberant sound with decaying amplitude is present.In addition to guard times,our algo-rithm compares the amplitude of the frequencies in arriv-ing signals.If during a time slot the ratio of amplitudes of two symbols is above a certain threshold,then the symbol with the lower amplitude is regarded as reverberation from neighboring time slots.

4.3.Synchronization

Our scheme requires mobile devices to issue a hello sig-nal at the beginning of an acoustic transmission to allow the server to synchronize to the transmission.We de?ne the hello signal in our system as a sinusoidal signal at a known frequency.We reserve the frequency of 2200Hz as the hello frequency.This value for the hello frequency is appropriate because:(1)it is within the optimal range of operation of speakers and microphones;(2)it maintains a guard band of 200Hz from the noisy indoor spectrum below 2000Hz;(3)it maintains suf?cient separation from the nearest symbol frequency,which is at 2489Hz;(4)its harmonic at 4400Hz maintains ample separation from nearby symbol frequen-

cies.

Samples

A m p l i t u d e

Figure 3.Signal shape of the received ID[1,2,3]:The amplitude values are normal-ized.

A related design choice for the hello signal is whether to send it just before the start of an ID,or to embed it within the ?rst symbol of an ID.Initial testing of these two cases has revealed that the server synchronizes more accurately to the signal when the hello signal is embedded in the ?rst symbol of the ID.Furthermore,synchronizing to the hello signal requires that it is at least a few milliseconds long,which implies that sending the hello signal prior to the ID incurs more delay in the decoding process.Consequently,our scheme includes the hello signal with the ?rst symbol of each ID.

5.Signal Reception and Decoding

5.1.ID Decoding

We begin by describing the decoding procedure for one symbol,and we later extend that process to decode se-quences of symbols that represent ID’s.Once the acoustic signal is received by the microphone,the signal must be de-coded in several steps with increasing granularity.First,the receiver discovers the beginning of the useful signal by syn-chronizing to the hello frequency.Once the server synchro-nizes to the mobile device’s signal,it partitions the acoustic signal into time slots with one symbol duration per time slot.The server then proceeds to discover the frequency content in each time slot.We use the Fast Fourier Transform (FFT)to transform the time domain acoustic signal in each time slot to the frequency domain.Let z be the received time do-main acoustic signal after sampling,and Z be the frequency domain representation of z .To derive the frequency content of Z ,we use the equation:

P zz =Z ×conj (Z )/N Z

(1)

where P zz is the power spectrum of the signal,and N Z is the number of points(samples)in the signal.

Next,we implicitly?lter the signal by examining the samples of P zz corresponding to the frequencies in Table1. In fact,for each frequency f in Table1,we examine the values of components of P zz in a small range of frequencies centered around f in order to account for hardware imper-fections that lead to frequency distortion.As we mentioned above,the frequencies in Table1are spaced far enough from each other so that a shift in one frequency does not crossover to the window of the neighboring frequency.If the value of P zz exceeds a threshold value at or close to a frequency f,then the server determines that f is present in the signal.Section6.1provides a more detailed discus-sion of setting the threshold.Once two or more frequencies that correspond to the same symbol are detected,the server determines that the symbol is present in the signal.By per-forming this quick check in subsequent time slots,the server can decode a sequence of digits to get the received ID. 5.2.Synchronization

Discovering the frequency content using the FFT method is applicable to individual time slots once the boundaries of each time slot are de?ned.De?ning the boundaries of time slots requires discovery of the beginning of the?rst time slot,which is achieved by synchronizing to the hello frequency.

Once the server captures the signal,it uses a sliding win-dow technique to discover the start of the signal.First,the server examines chunks of the signal that are equal in length to the embedded hello signal.Once it identi?es the sec-tion of the received signal with the highest correlation to the hello signal,the server then narrows the window of search for the hello signal to that section.Within that section,the server examines the signal samples in smaller chunks.By progressively shrinking the window size and the chunk size, the server eventually identi?es the sample that corresponds to the beginning of the hello signal,which is also the begin-ning of the?rst time slot in the ID(see section4.3).

Synchronization to the beginning of ID’s is even more valuable when several mobile users simultaneously perform location operations.Figure4illustrates the case of a mobile device sending its ID while the server is still in the process of receiving another user’s ID.Suppose the server receives the acoustic ID of mobile device A at time t,and the du-ration of a symbol is T.The receiver synchronizes to A’s transmission by detecting the start of the hello signal,and begins decoding the ID of A.Suppose also that at time t+a, where a is shorter than the duration of an ID,the receiver hears another transmission from user B.In Figure4,the acoustic ID of B arrives at the server during the second time slot of A’s transmission.While decoding the second sym-

Figure4.Synchronization

bol in A’s ID,the server detects that the hello frequency is present in the signal.As a result,the server synchronizes to the new transmission by performing a correlation check to synchronize to the hello signal of the new transmission,and initiates a separate thread to derive the frequency content of this transmission.Between time t+a and the end of A’s ID,the frequency content of the received signal is the com-bination of symbols sent by A and B.The FFT method de-tects up to six frequencies in each time slot,but the method cannot determine the source of each frequency.Thus,iden-tifying the received ID’s requires the following additional steps.

User A’s second symbol had started arriving at time t+T, so the server can identify the frequencies of A’s second sym-bol by checking the frequency content between t+T and t+a.The server can also determine B’s?rst digit by omit-ting A’s frequency content from the total frequency content received between time t+a and t+2T.To decode sub-sequent symbols of each ID,the server determines the fre-quency content of each time slot in the same way.

6.Experiments and Results

In order to validate our ability to detect multiple fre-quencies and symbols simultaneously through acoustic sig-nals,we conduct experiments in an of?ce of dimensions 18×9×9ft,using a typical PC microphone and speak-ers.In the experiments,the PC speakers send the frequen-cies corresponding to the encoded symbols through the air. The microphone then captures this audio signal,and once the sound card samples the data,we use Matlab[13]to per-form an FFT analysis to determine the frequency content of the captured signal.Table2contains the parameter values in our experiments.

Each experiment was performed on?ve different occa-sions,and the results re?ect the average of the?ve trials. The variance of results among separate trials of each exper-

Table2.Parameter Values

Parameter Value

Sampling Frequency(hz)11025

bits/sample8

Symbol duration(samples)1000

Symbol duration(ms)90.7

Guard time(samples)100

Guard time(ms)9.07

Room Dimensions(ft)18x9x9

iment was small,which suggests deterministic behavior of the system.

The?rst set of experiments explores the effects of dis-tance,frequency selection,and number of simultaneously emitted frequencies on our ability to detect transmitted fre-quencies.The second experiment set investigates the de-coding capability of the server to synchronize to and decode two asynchronous acoustic ID’s that overlap in time.

6.1.Frequency Detection

We?rst explore the effect that simultaneous symbol transmissions has on the server’s ability to detect individual symbols.A central feature of frequency detection ability in our scheme is the threshold value in the FFT method(see Section5.1).Setting the threshold value involves a tradeoff: a high threshold value insures that noise is not mistaken for a useful signal,but it also results in failure to detect some frequencies.On the other hand,setting the threshold too low allows detection of more frequencies,but it also in-creases the chance of false positives from noise.Thus,for each number of simultaneous symbols,we observe the min-imum threshold value within the FFT method that results in detection of all frequencies.We also use random combina-tions of the symbols(1,2,3)to avoid any biases introduced by frequency-speci?c behavior of the hardware used in our experiments.

Figure5shows the relative threshold value for the server to detect all the transmitted frequencies.For distances rang-ing from1ft to17ft,the server could detect all the fre-quencies in the cases of3,6and9simultaneously transmit-ted frequencies.However,the required threshold value for detecting each frequency using the FFT method(see sec-tion5.1)becomes lower as the number of simultaneous fre-quencies increase.

The detection threshold for6simultaneous frequencies (2simultaneous symbols)is about44%of the detection threshold of one symbol.Also,the threshold for9frequen-cies(3symbols)is36%of the1symbol threshold.The minor decrease in the detection threshold as the number of frequencies increases from6to9suggests that the received signal level degrades more slowly as more frequencies

are

0.2

0.4

0.6

0.8

1.2

1234567891011121314151617

Distance (ft)

Figure5.The normalized threshold value in

FFT method for full frequency detection as a

function of distance.

added.This feature is desirable for the scalability of the scheme.

The threshold and distance experiments revealed distinct channel responses for each combination of frequencies.To understand the channel frequency response for our hard-ware and to quantify these distinctions,we studied the re-quired threshold for different symbol combinations.Be-cause acoustic location systems do not necessarily use the same hardware or frequency set,the results in this study ap-ply only to our particular setting.These results also serve as a basis for the development of a general process that cal-ibrates thresholds for any hardware or frequency set.

Figure6provides a threshold comparison of all possible symbol combinations.Symbol(3)has the highest threshold among all symbol combinations.Symbol(2)and symbol (1)have respective relative thresholds of0.7and0.19.Two symbol combinations follow similar trend to the single sym-bol case,where the pair of symbols with the highest thresh-old is(2,3),and the pair with the lowest threshold is(1,2). The combination of the three symbols(1,2,3)has a thresh-old that is higher than both combinations(1)and(1,2).The results have2clear implications:(a)Symbol(1)contains frequencies with unfavorable channel response,and thus all combinations containing this symbol have low thresholds.

(b)Symbol(3)contains frequencies with favorable chan-nel response,and the presence of symbol(3)in a time slots improves channel response for other symbols(1and2).

Symbol Combinations

Figure 6.A comparison of the normalized

thresholds of frequency combinations.

6.2.Symbol Decoding

The next set of experiments investigates the server’s abil-ity to recognize two distinct ID’s arriving asynchronously. These experiments reveal that the main issue of correct de-coding is synchronization to the hello frequency through the correlation algorithm(see Section5.2).As the distance in-creases,the hello signal strength at the server is lower,and the correlation check at the server is more likely to yield an incorrect beginning of the signal.This degradation with dis-tance applies mainly to synchronization to the hello signal, and does not affect the detection of the frequency content as the experiments in Section6.1reveal.Thus,we explore the effect of both the hello frequency amplitude and the dis-tance on the server’s capability to decode2asynchronous symbols.

Figure7plots the percentage of symbols that are cor-rectly decoded by the server as the distance between the users and the sensors varies from1ft to17ft.Each of the three plots in Figure7is characterized by the ratio of the amplitude of the hello frequency to the amplitude of the symbol frequencies.For example,the plot for”Am-plitude=2”indicates that the hello frequency amplitude is double the amplitude of the symbol frequencies.

When the hello frequency amplitude is the same as the amplitude of all other frequencies,the server can synchro-nize to and decode only one of two acoustic ID’s arriving asynchronously.The decoding ability does not vary with distances of1to13ft,as the percentage of decoded ID’s remains around50%.Occasional peaks in the decoding

100

120

1234567891011121314151617

Distance (ft)

Figure7.Percentage of decoded symbols ver-

sus distance.

ability at distances of2ft and8ft are due to both hard-ware variations and multi-path re?ection effects within the experiment location.For distances of14ft or more,the decoding ability drops to40%because the hello frequency attenuation at these distances makes it more dif?cult to syn-chronize properly to the acoustic ID.On average,the server could decode49%of the transmitted ID’s for all distances within17ft.

If the hello frequency amplitude is doubled,the decoding ability at the server improves signi?cantly.First,doubling the hello frequency amplitude eliminates the effect of dis-tance on the decoding ability.As Figure7shows,the trend in decoding ability is constant for all distances within17 ft.A second observation is that the decoding ability varies between80%and100%in a seemingly random fashion. Again,this is attributed to multi-path effects and to varia-tions in the delay and response of the sound card according to the interrupt behavior and processing load at the host de-vice.Overall,the server could decode89%of acoustic ID’s.

Finally,the plot for”Amplitude=3”eliminates much of the variation in decoding ability that was observed for the previous case.Because the hello frequency amplitude is now triple that of symbol frequencies,the server can syn-chronize better to both acoustic ID’s at all distances within 17ft.The decoding ability of the server is perfect for most distances,and on average the server could decode98%of the transmitted acoustic ID’s.

7.Discussion and Conclusion

7.1.Background Noise

Some of the results were derived with music and other clutter noise sources in the background.Although exposure to noise sources was not systematic in all experiments,the ability of the server to decode symbols in the presence of clutter noise boosts our con?dence in the coding scheme.

7.2.Thresholds

Detection of symbol frequencies is not as dependent on the amplitude as synchronization to the hello signal,primar-ily because the server can detect symbol frequencies as the mobile device moves further from the receivers by lowering the detection threshold in the FFT method.Further research is required on algorithms to set these thresholds adaptively, especially when there are simultaneous transmissions from nodes that are both near and far from the microphones.Be-cause the server aggregates the received signals from sev-eral microphones,the effect of simultaneous transmissions from near and far nodes should be minimal.

7.3.Range

The receiver could detect all frequencies within a range of17ft,and the hello frequency ampli?cation enables the decoding of2asynchronous symbols at up to17ft.Future work will explore an extended range at distances up to the single frequency detection range(currently22ft)in larger deployment areas.The experiments will reveal the appli-cability of multi-frequency ID’s over larger ranges and in more diverse multi-path environments.

The importance of range in this system is offset by the fact that microphones are cheap.Thus,deploying a dense grid of ceiling-mounted microphones,each with a limited signal detection range,is cost-feasible.A dense grid of mi-crophones would enable the location system to capture any acoustic signal within the coverage area.

7.4.Calibration

Because the aerial acoustic channel includes the speak-ers,air,and microphone,channel behavior may be hardware-speci?c.Thus,calibration may be needed at some point prior to sending the acoustic ID’s.The results in Fig-ure6provide valuable insight into the hardware response to the current choice of frequencies.For example,the chan-nel has unfavorable response for the frequencies of symbol 1.Also,the channel has nonlinear behavior for various fre-quency combinations.Observing the speci?c responses of several speaker and microphone sets provides understand-ing on the trends in channel response,which can subse-quently be developed into a calibration process.

7.5.Security and Privacy

This acoustic location and identi?cation system provides a user with valuable context-speci?c information,but it also raises security and privacy issues.For example,department stores could track users’movements and behavior within the store,and use data mining techniques to market speci?c products to users.Therefore,the location system should include security measures to ensure the privacy of users.

In sum,we have presented a model of an acoustic lo-cation and identi?cation system that uses unique user ID’s based on multiple-frequency symbols.We have investigated the range of detection of1,2and3simultaneously transmit-ted symbols using3,6and9frequencies respectively.The results have yielded a range of at least17ft for up to9si-multaneous frequencies.Our other experiment has revealed that decoding asynchronous symbols at the server is highly dependent on synchronizing to the hello signal.To over-come this issue,we have proposed an increase in the ampli-tude of the hello signal by a factor of2or3.Once the server synchronizes to the signal,symbol decoding becomes inde-pendent of distance.As a result,the acoustic identi?cation scheme scales well for indoor location systems deployed in large rooms.

8.Acknowledgment

This work is part of a larger project that is funded by the UC MICRO Program.Raja Jurdak is in part supported by a grant from the Center for Pervasive Communications and Computing at UCI.The work of Pierre Baldi is in part supported by a Laurel Wilkening Faculty Innovation award, a Sun Microsystems award,and grants from NSF and NIH.

The authors would like to thank Atri Mandal and Amir Haghighat for their valuable insight and discussions on acoustic positioning and triangulation.The authors also thank the anonymous reviewers.

References

[1]J.Hightower and G.Borriello,“Location Systems for

Ubiquitous Computing,”Computer,V olume34,No.

8,pp57-66,August2001.

[2]R.Want et al.“The Active Badge Location System,”

ACM https://www.wendangku.net/doc/086936193.html,rmation Systems,1992,pp.91-102.

[3]A.Harter et al.“The Anatomy of a Context-Aware

Application,”In Proc.5th Ann.Intl Conf.Mobile Com-

puting and Networking(Mobicom99),ACM Press, 1999,pp.59-68.

[4]N.B.Priyantha,A.Chakraborty,and H.Balakrish-

nan.“The Cricket Location-Support System,”Proc.

6th Ann.Intl Conf.Mobile Computing and Networking (Mobicom00),ACM Press,2000,pp.32-43.

[5]Y.Fukuju,M.Minami,H.Morikawa,and T.Aoyama.

“DOLPHIN:An Autonomous Indoor Positioning Sys-tem in Ubiquitous Computing Environment,”In Pro-ceedings of IEEE Workshop on Software Technologies for Future Embedded Systems(WSTFES2003),2003, pp.53-56.

[6]P.Bahl and V.Padmanabhan.“RADAR:An In-

Building RF-Based User Location and Tracking Sys-tem,”In Proc.IEEE Infocom2000,IEEE CS Press 2000,pp.775-784.

[7]IEEE802.11WG.“Wireless Lan Medium Access

Control(MAC)and Physical Layer(PHY)Speci?ca-tions,”1999.Standard.

[8]C.V.Lopes and P.Aguiar.“Acoustic Modems for

Ubiquitous Computing,”IEEE Pervasive Computing, Mobile and Ubiquitous Systems,V ol.2,No.3,pp62-71,July-September2003.

[9]J.C.Chen,L.Yip,J.Elson,H.Wang,D.Maniezzo,

R.E.Hudson,K.Yao,and D.Estrin.In Proceedings of the IEEE,V ol.91,No.8,August2003.

[10]A.Madhavapeddy,D.Scott and R.Sharp.“Context-

Aware Computing with Sound,”In The Fifth Interna-tional Conference on Ubiquitous Computing,October 2003.

[11]“The Bluetooth Special Interest Group,”1999.

[12]F.A.Everest.The Master Handbook of Acoustics,

New York:McGraw-Hill,c2001.

[13]Matlab,https://www.wendangku.net/doc/086936193.html,/.

[整理]DCSSU-25操作手册

LOCK ON 中文讲解课程本章节主要讲解苏系战机起飞操作步骤飞机起飞步骤大体分为三个部分一：启动发动机二：滑出停机坪上跑道三：起飞首先是启动发动机的步骤，我们所实用的机型是 SU-25T ，只所以不是 SU-27 ，是因为 SU-27T 是 LOCK ON 中操作步骤比较复杂的机型，它的启动步骤包含其他机型的启动步骤，同时也有所复杂，所以我们用 SU-25T 来讲解飞机的启动步骤现在我们在 SU-25T 的座舱里可以看到飞机的所以仪表都处在关闭状态，因为主电源是关闭状态，我们通过 Shif+ “ R “键打开飞机的主电源，注意仪表开始有变化，观察气压高度计，我们通过 Shif+ “ - “和 Shif+

“ = “将气压高度计归零，这是 SU-25T 在 LOCK ON 中的标准操作步骤，至于为什么要将它归零，我也不知道！好，一切仪表正常，准备启动发动机，我们按 Ctrl+ “ L “键点亮翼灯，向塔台表明我们要启动发动机，确认节流阀收到低按键盘右侧（空格右边第 2 个键）文斗丝间和 HOME 键，让发动机启动，注意发动机的转速表开始变化，发动机已经启动， SU-25T 是双阀战斗机，目前版本的 LOCK ON 当中的所有可用机型都是双阀，所以我们可以单独的启动左发动机和右发动机，通过按 ATT GR+HOME 键来来启动左发动机，通过 Shif+HOME 键来启动右发动机，目前版本的 LOCK ON 中 SU-25T 以及 SU-25 的启动步骤是相同并且是最复杂的，其他大部分机型的启动步骤中，进入座舱后，电源本身是打开的，仪表也不需要矫正，直接启动发动机即可，现在我们回到 SU-25T 的座

员工入职资料表格汇总26387

入职资料登记表申请职位：入职日期：年月日基本资料姓名性别民族出生年月学历身份证号联系电话身高体重政治面貌是否有驾照及类型婚姻及生育状况户口所在地现居住地紧急联系人联系电话电子邮箱社会统筹情况□养老□医疗□失业□生育□工伤□公积金教育背景起止时间毕业院校专业学历/学位教育性质 □统招□函授□自考□其他 □统招□函授□自考□其他职业资格证书或其他相关证书：时间工作单位、职位离职原因工作经历证明人姓名及联系方式家庭关系姓名年龄工作单位职务联系方式主要

关系招聘渠道：□网络招聘□员工推荐（员工姓名）□人才市场□其他方式其他是否有朋友、亲戚在我公司工作？□是，请说明：□否其他说明事项：公司承诺：此资料将进入公司人才库严格保密，并仅作招聘使用。为全面了解您的优劣势，安排合适的岗位使您扬长避短，请您认真、完整、如实填写。本人承诺：本人授权公司向本人曾任职的公司、介绍人或咨询人查询所有记录，且申明以下提交的一切资料绝对真实，如有不实，可作为被公司辞退的理由，而公司无须做出任何赔偿。填表人确认签名：新员工录用工资确认表姓名录用部门录用岗位入职时间年月日发放日期每月20 日，遇节假日顺延薪酬标准执行①试用期工资：元/ 月；转正后元/月。第种；本岗位工资按公司规定暂实行足额发放。②实行年薪制 ③其他年薪元，试用期工资：元/ 月；转正后元/月；余额根据公司内部规定发放。 ①社会保险福利待遇③其它补助自年月开始缴纳社会保险。（备注说明：）补贴①：试用期发元/月；转正后发元/ 月。

补贴②：试用期发元/月；转正后发元/ 月。以上信息由人力资源部填写，填写人签字确认：确认日期：年月日人力资源部主管领导意见：确认日期：年月日领导签批确认：确认日期：年月日以下信息由新入职员工填写新入职员工身份证号码银行卡号开户行以上薪资及福利内容本人已经知晓，身份证号码、银行卡号、开户行信息由本人自已填写，并确保信息准确无误；因自已填写错误造成的损失由本人承担一切责任。同时本人承诺对自已的薪资福利绝对保密，如自已泄露愿接受公司的一切处理，直至解除劳动合同给予辞退。新入职员工确认签字：确认日期：年月日本表一式一份，仅供人事部与新录用员工核对薪资福利信息用，在经领导核定并经新录用员工签字备注说明确认后作为发放工资的依据。领导核定签批后原件留存于财务部门，人力资源部留存复印件。

su简单中文使用手册范本

CWP软件的安装与简单使用手册

CWP软件的安装一．在LINUX下建立用户CWP，在CWP下建立目录path，将源文件cwp.su.all.37.tar.Z 放二．到path目录下，并建立bin文件夹三．在CWP用户主目录下显示隐藏文件，修改.bash_profile 文件，在已有的export之后另起一行，分别添加 export CWPROOT=/home/CWP/path，再于 PATH=$PATH:$HOME/bin后添加：/home/CWP/path/bin:/home/CWP 退出保存四．从终端中分别输入 cd path zcat cwp.su.all.37.tar.Z | tar –xvf- … 待终端中反映完毕，分别输入 cd src make install make xtinstall make mglinstall make utils make xminstall make sfinstall 这期间可能有系统安装所等待的时间，不用急，但凡遇到yes/no，一路y下来即可。四．为了检查是否安装完毕，在终端中输入 Suplane > data.su Suxwigb < data.su & 若出现一个简单的图像，则成功！

CWP软件的简单说明一、文中涉及的命令全部以小写形式，均可在终端窗口下输入，以次获取自述帮助。先说几个命令：suplane和suxwigb，more。 suplane作用是产生一个简单的零偏移距su文件，suxwigb是一个典型的X—windows 绘制图形工具，如例子： suplane > data.su suxwigb < data.su more < data.su 比较全面的了解它们，请在终端中输入suplane , suxwigb ,more 。二、关于DEMOS的应用所有DEMOS必须把文件拷到用户根目录下，而后依照readme 文件中的执行顺序，在终端中输入文件名。注意目录下的文件变动。三、在执行DEMOS文件时，如果想清楚了解程序执行过程，请输入 more programname 由于水平有限，这里的谬误很多，希望大家能在偶尔翻看时，多多留心，发现并改正，衷心希望能和大家一起学习。

SketchUp教程[指南]

SketchUp教程[指南] SketchUp? 草图大师由于SketchUp直接面向癿是设计过程而不是渲染成品,不设计师用手工绘制构思草图癿过程徆相似,因此SketchUp癿目标是设计师做设计而不是制作员作图。 5、快捷键设置 L W O 线段漫游平行偏移 A ALT+` V 囿弧透明显示量觇器 N ALT+2 D 多边形消隐显示尺寸标注空格键 ALT+4 SHIFT+T 选择贴图显示三维文字 E F2 H 橡皮擏等觇透规规图平移 M F4 SHIFT+Z 移动前规图充满规图 S F6 F9 缩放左规图回到下个规图 J B K 路径跟随矩形绕轰旋转 Q C P 测量囿添加剖面 T F ALT+1 文字标注不觃则线段线框显示

Y X ALT+3 坐标轰油漆桶着色显示鼠标中键 G F3 规图旋转定丿组件顶规图 Z R F5 规图缩放旋转后规图 F8 U F7 恢复上个规图推拉右规图 I 相机位置，3，分割线段如果你在一条线段上开始画线,SketchUp会自动把厏来癿线段从交点处断开。例如,要把一条线分为两半,就从该线癿中点处画一条新癿线,再次选择厏来癿线段,你就会収现它被等分为两段了。，4，分割表面要分割一个表面,叧要画一条端点在表面周长上癿线段就可以了,

有时候,交叉线不能按你癿需要迕行分割。在打开轮廓线癿情冴下,所有不是表面周长一部份癿线都会显示为轳粗癿线。如果出现返样癿情冴,用直线工具在该线上描一条新癿线来迕行分割。SketchUp会重新分析你癿几何体幵重新整合返条线。，5，直线段癿精确绘制画线时,绘图窗口右下觇癿数值控制框中会以默认单位显示线段癿长度。此时可以输入数值。输入长度值输入一个新癿长度值,回车确定。如果你叧输入数字,SketchUp会使用当前文件癿单位设置。你也可以为输入癿数值指定单位,例如,英制癿(1’16”)戒者公制癿(3.652m) 。SketchUp会自动换算。输入三维坐标除了输入长度,SketchUp迓可以输入线段终点癿准确癿空间坐标。绝对坐标:你可以用中括号输入一组数字,表示以当前绘图坐标轰为基准癿绝对坐标,格式 [x, y, z]

Sketchup快速完全入门手册

Sketchup 快速完全入门手册焦志鹏1024 https://www.wendangku.net/doc/086936193.html,

写在前面：子曾经曰过：“工欲善其事，必先利其器”，两年前，我第一次接触su，中午收到别人从qq 上传来的su5.0，当时的感觉就是“这么小的软件”，当天下午了解了su 的大部分功能和基本用法，这时的想法是“果然是个小软件”…… 当时认为已经完全了解了su 的我在两年后的今天，却仍然不敢对任何人说：“学su？不会你找我！” 这就是su……

目录第一篇初识su 第二篇su 全局概述第三篇su 功能详解第四篇su 的材质和组件第五篇su 使用技巧

第一篇初识su 这一篇可以算作一个前言，讨论一些似乎看起来和su 没有关系，但实际上密不可分的问题。一、关于su 的适用范围以我所接触的人们来说，使用su 的人大概可以分为四类“建筑设计”“室内设计”“景观设计”“工业设计”，总的来说，每一类人里都有高手，也许每个人都在使用中有所感受，这里我只说说我的一家之言。建筑设计据我所知，su 被引入cad 行业，最早就是在“建筑设计”中被使用的，以我的观点看，su 作为建筑设计的模型制作工具，可以说是比较合适的。首先说说它的优点：第一，它的操作界面非常友好，相对于“三视图+相机视图”的模式，su 在人机交互界面上有着让人难以抗拒的友善感，在三维操作方面，也显得很灵活；第二，它的工具功能简单，操作方便，我一直认为，软件工具不怕多，就怕复杂，一个工具解决一个基本问题，多个工具组合解决复杂问题，这一点su 还是很不错的，在su 的工具栏里，只要单击就可以使用这个工具，单击不区分左右键，没有二级工具集，不弹出预设值对话框，一切都是顺序发生的，这一点对新手非常重要；第三，它的模型对简单体块的修改、查看非常方便，对建筑方案的前期推敲非常有利；第四，它的模型是单一文件，包含所有的贴图、块等元素，交换方便。其次我们也要面对它的缺点：su 最大的问题可以归结为两大类：程序内核本身先天的孱弱和与生俱来的功能的缺失，有这些问题本身并不可怕，真正让你感到不幸的是我要告诉你在可以预见的将来，我们对这两点无计可施，具体的问题会在后面陆续提及。说到具体的问题，有如下几点：第一，su 的性能和执行效率非常低下，我可以毫不客气的说它可以荣登我所用过的所有软件中“性能最差”的宝座；这是由它的内核决定的，这里我们不谈这个，第二，su 在设计上也存在很严重的问题，首先，su 的精度不够，su 的精度在长度上只能精确到小数点后1 位，既0.1，当你绘制一条0.11 单位长度的线时，你会发现在模型信息中这条直线的长度前添加了约等于符号“~”，也许你会说没有人会画这样的线，但是在有曲线相交的情况下，这种线是有可能存在的，su 在面积上只能精确到0.01，也就是说面积小于0.01 单位的面永远无法封面；第三su 设计上另一个缺失就是在su 中没有贝斯曲线，su 处理曲线只能将曲线转变为多段直线处理，换句话说su 不能处理曲线方程，比如在cad 里用spl 命令绘制的曲线，就是贝斯曲线，这种曲线导入su 后，将被转化为多段线，这样一些原本在cad 中交叉的线在su 导入的过程中就会断开，导致不能封面（其实cad 在处理spl 时也比较头疼，比如你可以试试，cad 里spl 无法延伸），而事实是更残酷的，实际上su 在处理任何曲线都是以多段线的形式处理的，因此cad 中的任何曲线，包括圆、圆弧、多段线中的弧都将无一幸免；第四，当你知道了su 连曲线都处理的如此狼狈，那么su 弱的可怜的曲面建模能力也就不足为奇了，很多人都在埋怨su 的曲面能力，但很少有人知道su 曲面能力弱是因为su 无法处理贝斯曲线，更没有人知道不能处理贝斯曲线实际上也是其精度不够的必然结果。说了这么多，总的看起来好像su 的缺点比优点要多，但是在实际应用中，我们只有在真正关注绝对精准的尺寸的施工图中才会遇到我上述各种缺点，因此，虽然它有很多不足，但仍然是当今非常热门的建模软件。室内设计用su 做室内设计的朋友们，可能会比用su 做建筑设计的朋友们更能感受我刚才说过的那些缺点，这是因为室内设计比建筑设计更关注细节，更容易遇到上述的那些不足，由于su 的性能不高，在场景复杂的情况下难以操作，由于su 的精度不够，一些细节的部分无法表达，由于曲线是多段线的形式，一些平滑曲面的效果必然会消耗大量的系统资源，曲面建模能力不足，也使很多设计中个性的亮点难以表现。总的来说，su 做室内设计，除了上手方

sketchup快捷键(中文版与英文版)

sketchup快捷键（中文版与英文版）中文编辑/撤销 Ctrl＋z 编辑/放弃选择 Ctrl＋t 编辑/辅助线/删除 Alt＋E 编辑/辅助线/显示 Shift＋ Q 编辑/辅助线/隐藏 Q 编辑/复制 Ctrl＋C 编辑/剪切 Ctrl＋X 编辑/全选 Ctrl＋A 编辑/群组 G 编辑/删除 Delete 编辑/显示/全部 Shift＋A 编辑/显示/上一次 Shift＋L 编辑/显示/选择物体 Shift＋H 编辑/隐藏 H 编辑/粘贴 Ctrl＋V 编辑/制作组建 Alt＋G 编辑/重复 Ctrl＋Y 编辑/将面翻转 Alt＋V 编辑/炸开/解除群组 Shift＋G 查看/工具栏/标准 Ctrl＋1 查看/工具栏/绘图 Ctrl＋2 查看/工具栏/视图 Ctrl＋3 查看/工具栏/图层 Shift＋W 查看/工具栏/相机 Ctrl＋4 查看/显示剖面 Alt＋，查看/显示剖切 Alt＋. 查看/虚显隐藏物体 Alt＋H 查看/页面/创建 Alt＋A 查看/页面/更新 Alt＋U 查看/页面/幻灯演示 Alt＋Space 查看/页面/删除 Alt＋D 查看/页面/上一页 pageup 查看/页面/下一页 pagedown 查看/页面/演示设置 Alt＋：查看/坐标轴 Alt＋Q 查看/X光模式 T 查看/阴影 Alt＋S 窗口/材质浏览器 Shift＋X 窗口/场景信息 Shift＋F1 窗口/图层 Shift＋E 窗口/系统属性 Shift＋P 窗口/页面设置 Alt＋L 窗口/阴影设置 Shift＋S 窗口/组建 Shift＋C 工具/材质 X 工具/测量/辅助线 Alt＋M 工具/尺寸标注 D 工具/量角器/辅助线 Alt＋P 工具/路径跟随 Alt＋F 工具/偏移 O 工具/剖面 Alt＋/ 工具/删除 E 工具/设置坐标轴 Y 工具/缩放 S 工具/推拉 U 工具/文字标注 Alt＋T 工具/旋转 Alt＋R 工具/选择 Space 工具/移动 M 绘制/多边形 P 绘制/矩形 R 绘制/徒手画 F 绘制/圆弧 A 绘制/圆形 C 绘制/直线 L 文件/保存 Ctrl＋S 文件/保存备份 Shift＋N 文件/打开 Ctrl＋O 文件/打印 Ctrl＋P 文件/导出/模型 Ctrl＋M

新员工入职资料

入职登记表个人基本信息姓名性别民族相片籍贯学历政治面貌婚姻状况□已婚有子女□已婚无子女□未婚□其他现住地址身份证号本人联系电话身份证住址提交资料 (复印件) □身份证□学历证□其他紧急联系人紧急联系人（请填写常住地的亲友）姓名关系工作单位和现住址紧急联系人电话家庭成员姓名关系工作单位家人联系电话健康信息利手：□左□右是否怀孕：□无□有是否曾被认定工伤或持有残疾人证明：□无□有：是否有重大疾病或家族病史：□无□有：是否从事过特别繁重体力劳动及有毒有害工种：□无□有：是否有职业病或慢性影响工作的疾病：□无□有：教育培训起止时间就读学校学历专业持有证书工作经历起止时间工作单位职位离职原因证明人/电话

单位意见：员工签名：年月日入职承诺书本人 , 男/女，学历：，身份证号码： , 1.本人在填写本《入职登记表》时，已保证自己符合国家法定的劳动年龄的标准，且与其他任何用人单位、机构、组织、团体无劳动关系；若违反前述承诺，导致用人单位被追究有关经济责任的，所有责任均由本人承担。 2.本人在填写《入职登记表》时，用人单位已如实告知工作内容、工作地点、工作条件、职业危害、安全生产状况、劳动报酬以及本人所需要了解的所有情况。 3.本人如有传染病、精神病或其他可能影响用人单位工作的病史，本人应以书面形式向用人单位说明。 4.本人承诺已与原单位解除劳动关系，且无仍然生效的保密协议、竞业限制协议。 5.本人填写的《入职登记表》所有信息真实有效，如有任何虚假，用人单位可按严重违反规章制度解除劳动合同，同时承担因此引起的所有责任。 6.本人承诺对用人单位相关信息（包括但不限于工资）承担保密责任。 7.如《入职登记表》中的信息有变化，本人有责任以书面形式向用人单位人事部门提交最新的信息。 8.本人承诺在试用期满经考核合格后即与用人单位签订劳动合同，本人将认真阅读并遵守各项规章制度。 9.本人承诺在用人单位任职期间不从事任何兼职行为。 10.本人所填写的通讯方式（包括地址、手机）均为有效，用人单位向任一通讯方式寄送或发出的文件或物品，如果发生收件人拒绝签收和已发送成功均视为送达。 11.本人承诺无被追究刑事责任记录，如有应以书面形式告知用人单位，如隐瞒事实真相，一切法律后果由本人承担。 12、本人承诺，遵守以下各项入职时的甲乙双方达成的约定：： ①新入职员工的试用期范围为一至三个月，用人单位将视试用期绩效考核确定转正期限。 ②员工在试用期需提前3日书面申请辞职，试用期满后离职必须提前一个月以书面形式向用人单位递交辞职申请，待用人单位批准后按指定日期办理工作交接，双方确认无误后方可离职。 ③凡未经批准或未办理正式离职交接手续而自行离职的员工，均视为自动离职，用人单位有权停发其工作期间的工资，作为离职后给公司带来损失的补偿。 ④自愿遵守公司规章制度，若在试用期3天内（含3天）不合适或自离的，不予工资结算。本人已充分了解上述资料的真实性是双方订立劳动合同的前提条件，本人填写的以上任何信息虚假

Su使用方式

Su：在su阵列中，在移动命令中输入“个数*”，则可以使往后阵列很多个。如果在二者之间阵列，则输入“个数/” 2 旋转背后的隐藏命令：选中要旋转的块，凭空点旋转（不在任何一个面上），以起始点为原点，任意一条线为起始边，选中的块会围绕起始点旋转，点击ctrl，可以环形复制（环形阵列）,如果在一定的角度中需要定角度等分，可以先复制（如果为180度内等分，则可以在选定块的180度的地方复制一个，再输入“个数/”则可以等分复制） 3 创建成一个组可以使几个个体各自互不干扰 4 放大缩小的时候在右下角的框中输入缩放比例就可以自动变化 5 拉伸的时候比例为-1时,表现为镜像 6 按住ctrl时缩放,表现为中心不动,向四周缩放 7 跟随路径：选择一个面，这个面为路径所围成的一个面，跟随路径，点击要跟随的那个面即可。 8, 卷尺工具隐藏：点住一点移动，可以移出辅助线；可以点住一点，在需要的方向移动一段可形成辅助线，点住辅助线可以将其复制 9 量角器可以画辅助线 10 轴：如果图形不与轴平行，在画矩形时不好画，可以使用轴，使得XYZ轴的方向改变，更改为你所需要的方向。回到原始的轴时，使用轴命令，在现有轴上点击右键“重设”就可以 11 尺寸：相当于cad中的标注 12 文本工具：相当于多重引线。从面上引出时引出的为该面的面积，可以更改标记 13 三维文本：形成的文字为立体文字，放置到面上时可以自动粘贴到该面上。一旦形成就为可变化的图形，可以在组件中更改。输错只能删除重新输入。 14 定位镜头：选择一个点将小人放到该点不松开选择视角，鼠标变为一双眼睛，可以四处观察，正面观察同理 15 漫游：按着鼠标进行漫游 16 视图--动画—添加场景，选择场景，在设置中，场景转换的时间表示每个场景所用的时间，场景延迟的时间表示在变换另一个场景的时候中间停留的时间。，如果场景延迟的时间为0则在到另一个场景时没有停顿。 17 设置动画的时候可以与定位镜头，漫游和正面观察相结合 18 导出动画时，可以设置导出时的参数，比如帧数率，大小 19 su可以导出cad文件，JPG文件等等 20 擦除键：使用时按着ctrl键可以使两个面变得光滑相交，如果想让光滑的面恢复，可以同时按ctrl和shift，再擦除。单独按shift时使用擦除，可以使线隐藏，而面不改变。要想使隐藏的线恢复，可在“编辑-隐藏-取消隐藏”中找回. 21 填充：调节密度在材质的对话框中的编辑里面。色轮边上第一个为还原颜色更改，第二个为匹配模型中对象的颜色，第三个为匹配屏幕中的颜色。（2）调节密度是在使用纹理图像输入窗口下面，不透明三个字上面的两个输入窗口中。（3）编辑时编辑的不是选中的面，而是编辑材质对话框中显示的那种材质。（4）按Alt键吸取材质。（5）创建成组的块在添加材质的时候可以一次全部添加材质，不创建成组的需要一个面一个面的添加材质。（6）还可以在推拉之前就添加材质，这样经过推拉之后得到的物块就是每个面都是已经添加了材质的了。

个人的SU使用经验

PDA建院平台约稿本来这篇文章的初衷是教一些SU技巧，但是想了很久以后决定还是不这样写了，因为其实所谓这些技巧光排列在这里没多少意义，我想既然是给大一大二的学弟学妹们的指导，更重要的应该是对SU全面的一个理解，并通过SU大致明白之后几年在建筑软件学习上的方向才是重点。所以我也无所谓跑题什么了，先讲一些比较关键的问题，并真心希望下面这些啰嗦的长篇大论能帮到你们。 1：SU到底难不难学简单来说，SU就是一个非常容易上手的建模软件。基本没有一个其他的软件能向SU 这样直观而且方便，因此只要你多使用它，必然能将它用的很好。有些同学会问，要不要报个什么班来学SU···如果真的这么想捐钱的话，捐给我好了··· 总之一句话：SU的技巧必须是建立在多使用的基础上，如果不去用只是光记住技巧那你是永远不可能用的好SU的（其他软件同理） 2：有RHINO和3DSmax这些强大的建模软件我们还有必要学SU么？我觉得这个要看个人喜好，一句话：哪个用着顺手用哪个。不过可以肯定的是SU足够应付你大学课程学习设计的所有要求（即使你的方案多么扭曲），而且毫无疑问也是效率最高的。RHINO能学也尽量学习，特别是GH。如果你毕业后能熟练使用这2个那你可以骄傲地说你在软件技术上已经击败了80%的同学了。至于3DSmax···我反正没说它不好用···但是真的不建议大家花时间去学，人家效果图公司赚点钱不容易，咱们就别抢人家饭碗了······ 不过建议大家多多尝试下一些新的软件，并搭配起来用，其实很多时候在SU里很麻烦的建模用其他软件可能一下就能做好，希望大家思路可以放宽。 3：SU对机器要求（买电脑的同学请注意）以下几点稍微注意：①CPU的核心频率是影响插件计算速度的关键，以及导图导CAD 的速度。单核或者多核对于SU没有任何影响（SU不支持多核模式），但是多核在渲染的时候比单核快比较多，因此尽量还是买多核的CPU②内存尽量大些，SU还是比较吃内存的，特别是你开着几个模型同时又开着CAD和photoshop的时候。因此从这个方面来讲系统最好是64位的（这样可以充分使用4G的内存）③显卡的显频比较关键，显频越高，SU的显示画质越好，还有就是显卡尽量买N卡，因为A卡对SU的支持经常出问题。 4：关于渲染··· SU自身的渲染能力其实已经是所有其他建模软件都比不上的，很多景观设计公司其实也就是SU直接导张大图就够用了，PS一下的话效果就更好了。此外SU还多众多风格可以选择，很多风格效果都相当棒，大家有空可以多试试看。所以我强烈建议大家一定要好好学习PS，PS能力好的话什么渲染器都是浮云。毕竟建筑设计还是看重设计逻辑和空间塑造。那种照片级的渲染图并不是我们奋斗的方向。因此不建议在渲染上面花过多心思，而且从华工老师评图的方向看，渲染水平并不能左右太多。还是那句话：人家效果图公司赚点钱真的不容易··· SU渲染器的选择还是有很多的，基本主流渲染器都是支持的，而且新出的渲染器也基本都会支持SU。下面详细说下各个渲染器的特点，大家自己根据情况选择。 ①VRAY：最老牌渲染器了，渲染效果很棒（前提是你会调整参数）。但是缺点很多：首先是VRAY的主要方向是支持3dsmax的（已经更新到2.0）而对SU的支持不是很在心（1.5都搞了1年还没出来），在SU里面经常会出现奇怪的问题，稳定性差。其次就是参数过于复杂，想学的话需要很多精力去啃那本《印象VRAY渲染》（想起来哥就为当初大二砸那么多时间而感到可惜）。而且就算你会调，当你出图的时候你会发现你根本没时间去调那

新员工入职流程通用版

新员工入职流程New employee orientation process 入职准备 1、人力中心向合格者发送《录用通知书》； 2、确认新员工报到日期,通知新员工在报到之前来公司明确报到需注意事项：所需资料、体检以及其他须知； 3、通知人事助理新员工报到日期，人事助理准备好新员工入职手续办理所需表单并负责依据《新员工入职通知单》内容落实各项工作： --用人部门负责安排办公位，申领电脑、电话； --行政办负责发放办公用品； --信息组负责开通邮箱、帐号、调试电脑设备等。 2入职报到 1、人力中心向新员工发放《新员工报到工作单》，并按要求办理入职手续： --员工填写《应聘登记表》，并交验各种证件：一寸免冠照片3张；身份证原件或户口复印件；学历、学位证书原件（学生提供学生证原件）；资历或资格证件原件；与原单位解除或终止劳动合同的证明；体检合格证明； --与员工签订劳动合同、保密协议、职位说明书； --建立员工档案、考勤卡； --介绍公司情况，引领新员工参观公司、介绍同事； --将新员工移交给用人部门； --OA网上发布加盟信息更新员工通讯录。 2、用人部门负责的工作 --负责安置座位，介绍并帮助熟悉工作环境； --制定专人作为新员工辅导员，介绍岗位职责和工作流程[1]。 3入职手续 1、填写《员工履历表》。 2、发放向新员工介绍公司情况及管理制度的《制度汇编》，使其具备基本公司工作知识，要求其通过公司内部网络了解进一步情况。 3、按照《新员工入职手续清单》逐项办理入职手续。 4、确认该员工调入人事档案的时间。 5、向新员工介绍管理层。 6、带新员工到部门，介绍给部门总经理。

SketchUp使用说明书

SketchUp?草图大师 1、设计相关软件分类与分析目前在设计行业普遍应用的CAD软件很多，主要有以下几种类型：第一种是AUTOCAD，及以其为平台编写的众多的专业软件。这种类型的特点是依赖于AUTOCAD本身的能力，而AUTOCAD由于其历史很长，为了照顾大量老用户的工作习惯，很难对其内核进行彻底的改造，只能进行缝缝补补的改进。因此，AUTOCAD固有的建模能力弱的特点和坐标系统不灵活的问题，越来越成为设计师与计算机进行实时交流的瓶颈。即使是专门编写的专业软件也大都着重于平、立、剖面图纸的绘制，对设计师在构思阶段灵活建模的需要基本难以满足。第二种是3DSMAX、MAYA、SOFTIMAGE等等具备多种建模能力及渲染能力的软件。这种类型软件的特点是虽然自身相对完善，但是其目标是“无所不能”和“尽量逼真”，因此其重点实际上并没有放到设计的过程上。即使是3DSVIZ这种号称是为设计师服务的软件，其实也是3DSMAX的简化版本而已，本质上都没有对设计过程进行重视。第三种是LIGHTSCAPE、MENTALRAY等等纯粹的渲染器，其重点是如何把其它软件建好的模型渲染得更加接近现实，当然就更不是关注设计过程的软件了。第四种是RIHNO这类软件，不具备逼真级别的渲染能力或者渲染能力很弱，其主要重点就是建模，尤其是复杂的模型。但是由于其面向的目标是工业产品造型设计，所以很不适合建筑设计师、室内设计师使用。目前在建筑设计、室内设计领域急需一种直接面向设计过程的专业软件。什么是设计过程呢？目前多数设计师无法直接在电脑里进行构思并及时与业主交流，只好以手绘草图为主，因为几乎所有软件的建模速度都跟不上设计师的思路。目前比较流行的工作模式是：设计师构思—勾画草图—向制作人员交待—建模人员建模—渲染人员渲染—设计师提出修改意见—修改—修改—最终出图，由于设计师能够直接控制的环节太少，必然会影响工作的准确性和效率。在这种情况下，我们欣喜地发现了直接面向设计过程的SKETCHUP。 2、软件公司简介 AtlastSoftware公司是美国著名的建筑设计软件开发商，公司最新推出的SketchUp建筑草图设计工具是一套令人耳目一新的设计工具，它给建筑师带来边构思边表现的体验，产品打破建筑师设计思想表现的束缚，快速形成建筑草图，创作建筑方案。SketchUp被建筑师称为最优秀的建筑草图工具，是建筑创作上的一大革命。 SketchUp是相当简便易学的强大工具，一些不熟悉电脑的建筑师可以很快的掌握它，它融合了铅笔画的优美与自然笔触，可以迅速地建构、显示、编辑三维建筑模型，同时可以导出透视图、DWG或DXF格式的2D向量文件等尺寸正确的平面图形。这是一套注重设计摸索过程的软件，世界上所有具规模的AEC（建筑工程）企业或大学几乎都已采用。建筑师在方案创作中使用CAD繁重的工作量可以被SketchUp的简洁、灵活与功能强大所代替，她带给建筑师的是一个专业的草图绘制工具，让建筑师更直接更方便的与业主和甲方交流，这些特性同样也适用于装潢设计师和户型设计师。 SketchUp是一套直接面向设计方案创作过程而不只是面向渲染成品或施工图纸的设计工具，其创作过程不仅能够充分表达设计师的思想而且完全满足与客户即时交流的需要，与设计师用手工绘制构思草图的过程很相似，同时其成品导入其它着色、后期、渲染软件可以继续形成照片级的商业效果图。是目前市面上为数不多的直接面向设计过程的设计工具，它使得设计师可以直接在电脑上进行十分直观的构思，随着构思的不断清晰，细节不断增加，最终形成的模型可以直接交给其它具备高级渲染能力的软件进行最终渲染。这样，设计师可以最大限度地减少机械重复劳动和控制设计成果的准确性。 3、软件特色 1、直接面向设计过程，使得设计师可以直接在电脑上进行十分直观的构思，随着构思的不断清晰，细节不断增加。这样，设计师可以最大限度地控制设计成果的准确性。 2、界面简洁，易学易用，命令极少，完全避免了像其它设计软件的复杂性。 3、直接针对建筑设计和室内设计，尤其是建筑设计，设计过程的任何阶段都可以作为直观的三维成品，甚至可以模拟手绘草图的效果，完全解决了及时与业主交流的问题。 4、在软件内可以为表面赋予材质、贴图，并且有2D、3D配景形成的图面效果类似于钢笔淡彩，使得设计过程

RClimDex中文用户手册

RClimDex (1.0) 极端气候指数计算软件用户手册张学斌 Feng Yang 加拿大环境部气候研究中心 2004年9月10日南京信息工程大学遥感学院陈昌春译注 2013．8

作者致谢 RClimDex 由Xuebin Zhang（张学斌）and Feng Yang（加拿大气象局气候研究部）开发与维护，最初的开发由加拿大国际发展办事处通过《加中气候协作项目，C5》资助。Lisa Alexander, Francis Zwiers, Byron Gleason, David Stephenson, Albert Klan Tank, Mark New, Lucie Vincent与Tom Peterson对R包的开发与测试作出了重要贡献。CCl/CLIVAR ETCCDMI的有关研讨会也对RClimDex的改进提供了宝贵的意见。. 译者的话原英文说明中所介绍的下载网址链接已无效，新网址包括 http://www.pcic.uvic.ca/tools-and-data/climdex https://www.wendangku.net/doc/086936193.html,/software.shtml RClimdex可计算极端气候指数27项，以下摘录来自一硕士论文《内蒙古地区极端气候事件时空变化及其与NDVI的相关性》（使用RClimdex软件）的15项指数名称翻译及一段简要说明。１.指数名称、解释、单位 FD0 霜日一年中日最低温＜0℃的日数天 SU25 夏日日数日最高气温＞25℃的日数天 GSL 作物生长期连续6 日>5℃或<5℃的时间跨度天 TN10p 冷夜日数日最低气温<10%分位值的日数天 TN90p 暖夜日数日最低气温>90%分位值的日数天 TX10p 冷昼日数日最高温<10%分位值的日数天 TX90p 暖昼日数日最高温>90%分位值的日数天 WSDI 热持续指数连续6 日最高温在90%分位值日数天 CSDI 冷持续指数连续6 日最低温在10%分位值日数天 RX5day 5 日最大降水量每月内连续五日的最大降水量 mm CDD 持续干燥指数日降水量<1mm 的最长连续日数天 CWD 持续湿润指数日降水量≥mm 的最大持续日数天 SDII 普通日降水强度降水量≥1mm 的总量与日数之比 mm R10 强降水日数每年日降水量＞=10mm 的总日数天 R95pTOT 强降水量 95%分位值强降水之和 mm ２.简要说明在应用RClimDex 处理数据之前，必须确保每个站点的数据以文本格式储存，并且储存的气象数据必须按照年、月、日、24 小时日降水量、日最高气温、日最低气温等顺序排列，各记录项之间通过空格将其隔开。由于研究的气象记录年限跨

Vray for SU 1.48.66 用户手册中文版(Kric翻译)

ASGVIS V-Ray for SketchUp Version1.48 New Features User Manual By Fernando Rentas Render by Adam Warner,Aura Studio 翻译：Kric V-Ray for SketchUp讨论群 44671563、95785049

V-Ray for SketchUp A Rendering Plug-in for Designers SketchUp users in all fields depend on V-Ray as a quick,easy and cost-efficient way to render their most cutting edge images. Developed with the Chaos Group,V-Ray for SketchUp works within Google’s SketchUp,one of the most popular 3D modeling tools available today. New features in version1.48include more shader types(great for metal textures),more than ten new procedural textures(including dirt),better material layers,faster parsing times,lights now have the ability to add IES profiles& more. Download the30day trial,or buy V-Ray for SketchUp now from https://www.wendangku.net/doc/086936193.html,!

Vray for SU 1.48.66 用户手册中文版

A SG VI S V-Ray for SketchUp Version1.48 New Features User Manual By Fernando Rentas R e n de r b y A d am W ar n e r,A ur a Stu di o 翻译：Kric V-Ray for SketchUp讨论群

SU系列使用手册

目录一、概况 2 二、工作原理3 三、外形结构及功能说明 5 四、规格表16 五、安装18 六、操作21 七、外接电池及保养22 八、异常状况处理23 九、通讯介面及说明26 十、UPS网络远程监控整体方案图27 十一、售后服务28 十二、装箱清单29 十三、附录（连接线径对照表）30

一、概况（一）、简介本系列是一款先进理想的在线式正弦波不间断电源供电系统，运用高频载波技术及IGBT功率器件，能精确控制UPS各种运转参数，为您的精密设备提供优质、可靠的交流电源。其应用范围广泛，从微电脑到大型计算机、通讯系统到工业自控设备都可以使用。本系列的在线式设计，在有市电时，会随时对市电电源不断追踪调整，在市电中断时，能不间断的从备用电池上提供储备电能输出。在过载或变流器失效异常的情形下，UPS 自动转换到旁路状态，由市电供电输出，若过载消失，会再自动恢复到正常状态供给设备使用。转换时间均在 4 毫秒内，使您的设备完全不受影响，实现全自动不间断供电，提高了您电脑及设备的利用率及耐用性。（二）、注意事项为避免机器损坏或伤害使用者，使用前请详细阅读及遵守以下事项： ●请避免将机器置于潮湿或过热的环境中。 ●请避免将电源线连接于不符规定的插座中。 ●请将机器置于空气流通的场所，并保持散热孔的通风。 ●请勿随意打开机器外盖，以避免触及高压危险。如用户自行拆机，本公司将不负责维修责任。 ●请勿倾倒任何液体、物体于UPS内部。 ●外接电池及电池组应远离火源及热源。 ●如有任何不良现象情形，请立即将市电插头拔掉或关掉开关，并联系代理商专业人员及维修中心。

二、工作原理 (一)、UPS系统架构方框图 (图一) UPS系统架构方框图 (二)、UPS正常运转时的运作方法：当UPS正常运转时，如图二（黑色标色部分所示）：一路经由充电器对电池组充电，保持电池电压于满电位，另一路则由整流器整流后经逆变器转成纯净的交流正弦波电源，再由静态开关转换，送至用户设备使用。 (图二)正常运转时之运作方法

SU-2000用户手册(3月20日)

SU-2000火焰检测器用户手册

目录 1 介绍 (1) 1.1 产品开箱确认 (1) 1.2 声明 (1) 2 描述 (2) 3 主要部件 (3) 3.1 挠性光纤组件 (6) 3.2 观测管组件 (6) 3.3 安装管组件 (7) 3.4 冷却风软管 (7) 3.5 手动球阀 (7) 3.6 火检探头 (7) 3.7 电缆组件 (8) 3.8 电源组件 (8) 3.9 联网软件 (8) 4 安装 (9) 4.1 观测管组件的安装 (9) 4.2 挠性光纤组件的安装 (10) 4.2.1 外导管组件的安装 (10) 4.2.2 内导管组件的安装 (11) 4.3 安装管组件的安装 (11) 4.4 冷却风软管的安装 (11) 4.5 火检探头的安装 (11) 4.6 电源组件的安装 (11) 4.7 电缆组件就地接线盒的安装 (11) 4.8 电气连接 (12) 5 调试 (13) 5.1 冷态调试 (13) 5.2 热态调试 (14) 5.2.1 热态调试-油火检调试 (14) 5.2.2 热态调试-煤火检调试 (15) 6 操作 (17) 6.1 自动选择鉴别频率 (17) 6.2 手动选择鉴别频率 (18)

7 常见问题处理 (19) 8 维护 (21) 8.1 光纤 (21) 8.2 内导管组件 (22) 8.3 外导管组件 (22) 8.4 观测管组件 (22) 8.5 冷却风软管 (22) 8.6 火检探头 (23) 8.7 电缆组件 (23) 9 仓储 (24) 10 产品返修 (25) 11 备件采购 (25) 12 表单模板 (26) 12.1 RMA维修联络单模板 (26) 12.2 备件询价单模板 (27) 其他附图 ?挠性光纤组件 ?观测管组件 ?电源分配回路图 ?火检系统接线原理图 ?电源柜外形尺寸示意图 ?就地接线盒及接线示意图 ?火检系统联网图

(完整版)HR-9新员工入职办理流程及员工档案资料清单

一、新员工入职办理流程第1步：新员工报到，人事专员接待（谁招聘谁办理）。第2步：人事专员先收集：员工应提供的资料，核对并复印。第3步：人事专员提供我方应准备的资料：劳动合同、保密协议、告知函、人事制度等，给新员工学习，并讲解，时间控制在15分钟以内。第4步：告知其薪酬为面试时的约定，一周内由财务来核定其薪酬如何切分。第5步：员工签订相关资料后，用手机钉钉通知IT部，开通邮箱/钉钉等权限，开通后由IT部发到其钉钉上。第6步：人事讲解公司基本考勤制度。第7步：人事介绍公司/部门/岗位基本情况。第8步：人事协助员工安装钉钉，并讲解基本的钉钉使用技能。第9步：引领员工，认识主要部门负责人，并熟悉公司环境。第10步：引导员工：到IT部领取电脑设备，到行政部领取办公用品。第11步：引导员工到部门负责人处报到，并安排好办公桌。第12步：邮箱发送新员工报道通告，员工档案转交。第13步：员工入职报到流程结束。二、员工基本档案资料员工提供资料： 1.身份证原件、复印件，公司核对原件，留存复印件； 2.学历证书原件、复印件、学位证书原件、复印件（一般为最高学历），公司核对原件，留存复印件； 3.专业技术职称证书原件、职业资格证书原件、上岗证书原件，公司核对原件，留存复印件； 4.上家公司离职证明（原件）； 5.上家公司劳动合同，公司核对原件，留存复印件； 6.体检报告：最近三个月内、三甲医院体检证明原件； 7.银行卡复印件（标配光大银行卡，入职一个星期后收取）； 8.就业失业登记证（上海户籍员工）；

9.公积金账号；人事专员准备资料： 10.个人简历； 11.员工求职/入职信息登记表； 12.面试评估表（给入职者看薪资）； 13.劳动合同（需签收）； 14.保密协议（需签收）； 15.用人单位基本信息告知函。