Motion Distillation for Pedestrian Surveillance
Motion Distillation for Pedestrian Surveillance
M. Sugrue and E. R. Davies
Royal Holloway, University of London
{m.sugrue, e.r.davies} @ https://www.wendangku.net/doc/6a11013146.html,
Abstract
Motion detection in video is a fundamental step in all object tracking tasks; however, it is often incorrectly treated as a solved problem. The most common approaches are statistical background modelling and condensation. However, these methods have certain inefficiencies in their use of motion information. The output of background modelling is change detection rather than true motion detection, and is highly susceptible to non-motion pixel change due to noise and lighting changes. Condensation techniques use only spatial or ‘foreground’ information and must be initialised with target appearance information.
Here we present a new wavelet-based method for true motion detection which incorporates spatial and temporal information on an equal footing. The algorithm is tested using noisy outdoor tracking scenarios, can self-initialise and track arbitrary objects, and is shown to be inherently both fast and robust.
1. Introduction
Much recent research into motion detection and tracking tends to view video as a sequence of frames, ignoring the wealth of temporal information which is directly available in volumetric (x, y, t) representations. Motion tracking schemes fall into two broad categories, those based on foreground tracking (principally particle filtering or condensation) and those which use an initial background modelling/subtraction step (such as median filtering, Gaussian Mixture Modelling, etc) followed by an ‘inertial’ prediction model such as a Kalman filter. Both schemes, however, have inherent inefficiencies in how they deal with motion information. Particle filtering concentrates on searching for a predefined appearance model of the target, examining only foreground information and, except for the recording of previous foreground positions for use in prediction, essentially ignoring data in the temporal domain. The result is an inefficient need to ‘over-search’, using a huge number of multiple hypotheses, and requiring to check them all in each frame. Background modelling, alternatively, tries to reduce the search area by eliminating stationary ‘background’ features. This requires computing statistical likelihoods for each pixel in the frame. The reduction of the search area eases the strain on the subsequent prediction step, allowing less complex models to be used, such as the Kalman filter.
Background modelling schemes face two inherent and antagonistic problems, called the Stationary Background problem and the Transient Background problem [1]. First, the background model must reflect the stationary part of the scene to allow accurate segmentation of moving objects, required by frame-to-frame correspondence-based tracking [4]. This problem requires a low difference threshold in the subtraction stage and a wide temporal window so that slowly moving objects do not merge with the background. Second, the background model must update to appearance changes in the scene, such as changed lighting conditions, requiring narrower temporal windows. No compromise gives perfect results, a common failure being partial segmentation when the ‘object depth’ (object overlap in previous frames, equal to length/speed) is greater than half the temporal window. The statistical methods most common for this, (such as temporal median, Gaussian mixture models [8], etc.) are essentially pixel-based one-dimensional (time only) processes and could be more correctly described as ‘change detection’ algorithms, and thus are highly susceptible to non-motion pixel change due to noise and lighting variation, as well as such practical surveillance problems such as camera shake.
Even in the ideal case such approaches provide only binary detection information, defining which areas of the current frame are in motion and which are not. Target understanding must be built on top of this level: most applications attempt to model target behaviour based solely on the path of the object’s centroid [5].
This approach was developed for, and works well with, traffic monitoring. Cars are rigid objects that largely translate across the video frame. However, when applied to pedestrian tracking much useful information is lost. Pedestrians are non-rigid objects that change appearance through the very act of walking. Further, when the end goal of the system is not merely tracking but some higher level behaviour understanding, binary segmentation discards much useful information. Looking at motion detail inside the segmented motion ‘silhouette’ may
reveal interesting information about pedestrians, but little about cars. Studies of CCTV industry employees have shown that humans recognise behaviour based on considerations such as posture or limb or body movements, and rarely if ever on object path alone [10].
Our approach has attempted to use volumetric processing methods to extract rich, non-binary motion information, a goal we have termed ‘motion distillation’. Good pre-processing which passes quality motion information to higher levels greatly aids unusual behaviour detection.
Volumetric video processing algorithms have infrequently been mentioned in the literature. One example where xt ‘slices’ of volumetric video were analysed for the distinctive periodic pattern of a pedestrian’s foot motion appears in [6]. In neuro-physiology, the MT (middle temporal) and V2 areas of the brain contain local speed-specific and motion direction-specific cells. There have been several attempts to implement this in hardware, such as a chip [3] which measures speed using the time it takes the edge peak to travel from one pixel to its neighbour. Delbrück [2] uses a time delay circuit to compare one frame with the next. The recent work of Irani has focused on the many uses for spatio-temporal texture analysis, including video completion [11] and behaviour analysis [7].
The remainder of this paper is organised as follows. Section 2 describes our method for motion distillation including details of computational load. Section 3 presents results of two tests of our method. Each test uses three videos with differing noise properties. The first is a quantitative measure of motion detection; the second test is a qualitative segmentation comparison of our method with two widely used background modelling methods – the temporal median filter and the Gaussian Mixture Model. The paper concludes with an appraisal of the approach and some indications for further research.
2. Method
The aim of this research is to develop a video pre-processing scheme which allows for direct and robust object detection and tracking, as well as a useful input for a motion-based scheme for object categorisation and understanding. We approach this problem using the uncommon concepts of volumetric video and motion edges.
Volumetric video involves stacking frames into a long column and processing it using ‘bulk’ 3D operators. Motion can be distinguished from change or noise by its spatial coherence. As will be shown below, this means that motion detection may be performed using a temporal edge detector in the ‘temporal planes’ of the image – viz. xt or yt. Edges which are parallel to the t axis belong to objects which are stationary with respect to the video frame. Edges with a component perpendicular to the t axis are in motion. The authors’ previous work [9] used a horizontal (xt) 3 × 3 Sobel edge detector to enhance motion edges; a suitable threshold was then employed to produce a binary motion map. Here we extend these ideas using spatio-temporal wavelet decomposition to achieve what we call ‘motion distillation’.
Objects are tracked using a dual channel form–motion algorithm whereby objects in motion are tracked solely by their motion. Ambiguities due to non-detection of objects that stop moving are resolved by reference to image data
in the form channel, using template matching.
(a) Original Data
(b) 1st Scaling
(c) 1st Wavelet
(d) 2nd Scaling (e) 2nd Wavelet
Fig. 1. Example of Haar wavelet decomposition
2.1. Motion channel computation
Wavelet Transforms are formed of two functions, the Scaling function (low-pass filter) which behaves like an averaging function, and a Wavelet function (high-pass filter) which is essentially an edge detector. The functioning of the 3D filters we use can best be explained by analogy to the simplest 1D example, the Haar wavelet, where the Scaling function is the set {1, 1} and the Wavelet function is {1,
–1}. For multiscale decomposition of a linear input signal, the scaling and wavelet functions are successively convolved with the signal.
As can be seen in Fig. 1, the wavelet function highlights discontinuities at its scale of operation, while the scaling function outputs the data ‘trend’ at its scale. Wavelet decomposition can easily be extended to higher
dimensions using the tensor products of 1D wavelets. There are four functions for processing planar signals such as images: these comprise the scaling function and three edge detecting wavelet functions for horizontal, vertical and diagonally orientated edges. For the 3D (spatio-temporal) case there are eight filters. Of these, this project uses only two, the 3D scaling function (Fig. 2a) and the 3D horizontal wavelet function (or temporal edge detector, TED) (Fig. 2b), which we use to enhance motion edges. The TED is a motion detector, rather than a change detector, because it incorporates spatial as well as temporal information in a coherent manner.
(a) 3D Scaling function (b) 3D Wavelet function, TED
Fig. 2. Diagram showing the two Haar functions used. White represents 1 and black –1.
The amount of motion noise in a video can be assessed by examining how pixel values vary over time about their mean value. When the amount of motion in a video is small with respect to the background, a common case in surveillance videos, a histogram of pixel variances will show a large peak, as in Fig. 3. The background modelling paradigm relies on the ability to distinguish pixel change due to moving objects from usual pixel variance due to noise. As can be seen from Fig. 3, this is a difficult task when following this approach. Using one-dimensional statistics, pixel variance and motion are not easily separated. Figure 4 shows the result of processing the same video with a TED. Moving objects form edges in the ‘volume’ of the video. Pixels due to motion are enhanced by the TED and returned with high values, while low value pixels represent non-motion background.
Haar wavelet function behaviour can be seen as suppression of static components of a signal. Higher order wavelets, such as the bi-orthogonal D5/7 wavelet used in JPEG 2000 standard, suppress higher order polynomials – good for data compression, but not advantageous for simple edge detection.
When used spatio-temporally on video, the static component is the background and so the wavelet function will highlight motion at its scale of operation. However, noise at the filter’s scale will also be detected. At higher decomposition levels, video noise (typically random and without structure) is scaled out. The filters also show precise speed sensitivities at different scales. The first wavelet convolution enhances edges which are moving at a speed of greater than one pixel per frame. A scaling convolution followed by a further wavelet pass returns edges moving at greater than 0.5 pixels per frame, but with half the resolution. Further decomposition steps successively double the speed sensitivity while halving the resolution. For slow motion detection at decomposition level D, each band shows motion greater than 2–D pixels per frame.
Fig. 3. Pixel Variance of 20 frames of video containing moving objects
Fig. 4. Motion Distillation output. Low pixel values due to non-moving background, high pixel values due to motion.
The reduction of resolution can be viewed as a natural result of uncertainty. If an edge moves 2 pixels from one frame to the next, its location in the motion output cannot be given at greater precision than ±1 pixel, so one datapoint (pixel) is output. If greater precision is required, the lower decomposition outputs can be reintegrated as in the results section below. This inherent smoothing effect is seen in Fig. 7 (Frame 5185).
Figure 5 shows a flow diagram for the entire system. The motion distillation output is searched for blobs (connected regions) and these are grouped together into tracked objects frame by frame. The final motion-based understanding component is discussed in Section 3.3.
Fig. 5. Overall flow diagram of the approach. It is intended that the final object tracking output be fed to a motion-based scene understanding stage, but in this paper the concentration is on effective unaided motion extraction.
2.2. Computational load
Frame rate comparisons of techniques are problematic due to differences in implementation, input and equipment. A more precise measurement of algorithm speed is a calculation of required operations per pixel, which we attempt below. However, we can report that our implementation of the Motion Distillation scheme runs at 62 fps on a P4 machine, while median filtering and Gaussian Mixture Model (GMM) run approximately 10 and 80 times slower. Input frame size is 720 × 576.
The computational requirement of the temporal median filter method of background modelling is ~256 operations per pixel (because of the need to analyse the intensity histogram). The popular GMM technique is more expensive, requiring a lengthy initialisation step, followed by evaluation of an exponential function for each Gaussian and each pixel.
For Motion Distillation, the concept of temporal window is replaced by that of decomposition level, D. The number of starting frames for this decomposition level is 2D. On the first iteration, two frames are convolved with two filters – the scaling function and the wavelet function – to produce one scaled frame and one motion frame (of size n/2 ×n/2, where n is the frame width). The next stage repeats this for two scaled frames from stage one, resulting in output frames of size n/4 ×n/4. The third decomposition stage uses two second level scaled frames and produces frames of size n/8 ×n/8.
The total number of pixels in the system is given by:
∑
=
=
D
i
i
D
i
n
N
2
2
2
)
2
(
(1)
For the 3D Haar Wavelet Transform, the number of
operations required to decompose the signal to level D is
given by equation (1) but with the expression summed to
D – 1. The number of operations per pixel is:
∑
∑
?
=
?
?
==
1
3
2
1
2
2
2
2
2
)
2
(
D
i
i
D
D
i
i
D
i
n
n
(2) the minimum value of which is 1 operation per pixel
when D = 1 and the maximum is found using the limit as
D goes to infinity:
(3)
14
.1
7/8
2
lim
1
3=
=
??
?
?
??
?
?∑?
=
?
∞
→
D
i
i
D
This measure of computational load, at < 1.14
operations per pixel per filter, is close to the minimum
possible, and is a major speed improvement when
compared with that of the median filter or other methods.
3. Results
In this section we present data showing the raw output of
the motion channel and basic tracking results for our
system. Quantitative results were computed for three test
videos. Object detection rates for each video were
compared with the manually established ground truth. In
Table 1, TB stands for 'True Blobs', which is the true
number of moving objects in each frame. FP and FN
stand for False Positive and False Negative. In test videos
1 and 2, all false positives were observed to be due to
moving tree branches and all real moving objects were
detected. In the CAVIAR video a small number of false
positives were caused by shadows and a poster moving in
air currents. The 24 false negatives were due to a single
person moving slowly in a dark region of the video
(though there is clearly some argument whether these
false negatives should be counted as true positives).
Table 1. Performance of motion channel with three videos.
#
Frames TB FP FN Precision
Video 1 538 378 2 0 99.24%
Video 2 5458 2712 138 0 94.91%
CAVIAR 550
938
20
24
95.30%
For the segmentation tests we have chosen to compare our method with two background model and subtraction methods because they, like our method, have no requirement for a priori target information and they output a segmentation map of ‘moving’ pixels. We have chosen the temporal median filter because it has comparable algorithmic complexity to our own method, it may be set up to operate over a small number of frames and it has a simple bootstrap. The GMM approach is chosen as it is an accepted, particularly widely used background modelling method. Here both the median filter and the new method are computed over eight frames (the new method is computed to the third decomposition level, for which the number of frames is 23), and both also have an 8-frame bootstrap. GMM requires a larger number of frames in order for the Gaussians to stabilise on a particular distribution, and in our implementation we use a 20-frame bootstrap. The output from all three methods is presented without any subsequent morphological or noise reduction steps. It is also intended, that by contrasting our new method with two common accepted background modelling techniques, we may demonstrate that their weaknesses are general to the paradigm, rather than specific to any one implementation. 3.1. Tracking results
Simple tracking requires motion object detection and the extent of the object to be determined with good accuracy though not necessarily high precision (we shall clarify this point below). Here we show the tracking output for our technique. One important difference here is that with background modelling the motion mask must be of the same size as the input frame and it must be searched for objects using connectivity analysis. Through spatio-temporal scaling, the new method produces an output result 22D times smaller than the frame area (in this case, 64 times smaller) and thus a greatly reduced object search time.
The results shown in Fig. 6 were achieved based on motion alone, without the need for a predictive device such as a Kalman filter. (See the discussion in Section 4 relating to situations when occlusions occur.) The three tracks produced by the system are: (1) the pedestrian track, which shows the path of the target centroid including the slight ‘bobbing’ motion due to walking; (2) the track of the pedestrian’s reflection in the window; and (3) the hardly noticeable motion of a distant pedestrian seen through the branches of a tree (see the top left hand corner of figure 6). It is interesting to note that the method can clearly distinguish the motion of this pedestrian from the large amount of image noise at this point in the trees, even though the object is only a few pixels in size. The detail of this result is discussed in Fig. 7 below.
3.2. Segmentation comparison
Here we examine more closely the inner workings of the method. If precise segmentation of motion is required, the result of the third level of decomposition can be used as a mask for previous output, and the edge of the silhouette can be refined using the lower level decomposition outputs. Below is a comparison of results from several videos with different lighting and noise properties. The results of the new method have been thresholded to produce a silhouette comparable to background subtraction. However, the true output of the method is not binary: this is discussed in the next section.
We present here segmentation results from three video segments. So as to compare our method to background modelling techniques whose performances are highly dependent on noise levels, the videos were chosen for their differing noise levels. Video noise may be characterised by the median value of pixel variance over time (As discussed in Section 2.1 and Fig. 3). The test videos also display a range of pedestrian motion speeds and behaviours, and provide an important test of the windowing problem discussed in Section 1.
The first case presented here (Fig. 7, Video 1) is a simple motion segmentation task of an outdoor scene with diffuse lighting. The median pixel intensity variance is 1.65. The median filter results clearly show the difficulties of that method. The target is incorrectly segmented with leading and trailing edges separated. This is due to the slow speed of the target with respect to the temporal window size of 8 frames; the middle of the target has been absorbed into the background model. Both the GMM and new method give more accurate results.
Video 2 in Fig. 7 shows a variety of target pedestrian behaviours. The pixel variance is 3.05. In frames 665 and 767, the target is walking directly across the frame. There are slight shadows which interfere with segmentation and more random image noise than Video 1. This noise shows clearly in both median and GMM background subtraction results because these methods behave like a change detector. Pixels are segmented if the contrast with the background model is above some threshold. (In median filtering this threshold is global, in GMM it is pixelwise and adaptive.) Morphological closing (which would have to be anisotropic, and would to a fair extent be an ad hoc measure) could improve the background subtraction results, but this somewhat expensive step is unnecessary for the new method. (However, it is commonly necessary for many implementations, such as [8].) The random
Fig. 6. Object tracks from a 120-frame video using the new method. The three tracks shown are the pedestrian, his reflection in a window, and (top left) the movements of another pedestrian, seen through the branches of a tree.
Video 1
Video 2, Frame 665
Frame 767
Frame 5133
Frame 5185
Frames
Output Median Filter
Output GMM
Output New method
(Detail of Frame 767)
Fig. 7. Showing output from two videos. Video 1 is low noise outdoor scene. Video 2 is high noise, shows variety of pedestrian behaviours. The bottom row shows an enlarged section of Frame 767 with outputs.
structureless nature of the video noise means that the TED of the new method reacts less strongly, and is automatically removed by the scaling process. In this frame there are also two other small regions of motion (as
discussed in relation to Fig. 6). Median filtering detects only one of these, while GMM catches both, but in neither of these cases is a strong signal obtained. However neither method is capable of cleanly distinguishing these objects
from noise and it is likely that subsequent noise reduction steps will remove them entirely. The new method clearly detects both small moving regions while robustly eliminating all noise.
CAVIAR
Video
Median
Filter
GMM
New
Method
Fig. 8. CAVIAR database video. Half PAL format.
Frame 767 shows an important strength of the new method. Here the target has passed partly into the shadow of a tree. The background models fail to properly segment the legs because their intensity is too similar to the background. However, the motion of the legs, and thus the response of the TED is still sufficient to detect the full object and the new method segments the target correctly, as can be seen in the detail inset.
Fig. 7, Frame 5133, is of a target pedestrian walking slowly towards the camera, a case rarely dealt with in the background modelling literature. Because of the slow motion relative to the image frame, the centre of the target becomes absorbed into the background (to a large degree in the median filter, and less so in GMM), and the background subtraction results show large gaps. In Frame 5185 the pedestrian is waving his arms, this is discussed further in the future work section below. The new method results for Frame 5185 show slight ‘smearing’ of the arms due to their quick motion, which is a characteristic by-product of this method. (It must be emphasised that motion analysis is necessarily carried out over time, and thus refers to a range of positions: the complete picture at any moment can therefore only be
ascertained by combining form and motion information.) Again, the new method also detects a small moving object in the top left corner which is completely missed by both the median and GMM methods. Figure 8 is of the “fights & runs away 1” sequence from the CAVIAR database. This video shows the highest degree of noise of these examples with pixel variance as high as 10 in the brightly lit bottom left quadrant of the video. Again the new method has a cleaner response because this noise is random, and thus changing but not moving.
In all cases the new method demonstrates a far greater robustness to noise than either median filtering or GMM.
3.3. Future work: target understanding and analysis
The results presented above serve to compare the ability of the new method to detect motion from non-motion in a Boolean manner. However, the natural output of this method is not binary as with background modelling, but contains a wealth of sub-object level motion information which has been thresholded to produce the above comparisons. Our current research is on how best to utilise this extra information to achieve target understanding and behaviour analysis. We present some example output below (Figs. 9 and 10).
Fig. 9. Target motion field history superimposed on the current frame.
Fig. 10. Frame 5185 from Fig. 7, with insert showing instantaneous motion field of target.
Figure 9 shows a representation of the motion history of an object. The intensity of the red blocks indicates the maximum speed of that part of the object as it passed that point in the frame. It can be seen that the cyclical walking pattern of the pedestrian is revealed. This data can be used to both identify the nature of the moving object and its behaviour: it is a pedestrian if it has fast cyclical motion at the bottom; the pedestrian is walking and behaving normally if there are no other unusual motion patterns. Figure 10 shows an example of unusual behaviour. The figure shows a pedestrian waving his arms, with an inset showing the instantaneous motion field. Quickly moving arms are shown in red and slowly moving body is shown in blue. The pedestrian’s right knee is being brought forward at an intermediate speed and so is displayed as purple. This behaviour could be identified as unusual by comparing the motion profile of this figure with a database of common motion profiles.
4. Discussion
The key advantages of this preprocessing scheme over background modelling-based systems are its speed, simplicity and robustness. The system achieves tracking by directly using the motion field output, without the need for any predictive elements, such as a Kalman filter. In cases of ambiguity such as the merging of two objects, or occlusions, a higher level processing stage would be required to connect discontinuous tracks. We are currently investigating schemes based on object appearance and ordered sets.
This new method is a credible alternative to background modelling in every case we have tested. It sidesteps the temporal window problems discussed in Section 1, is far less computationally expensive, and provides a strong foundation of motion data for higher level analysis and understanding. However, in this study we have only dealt with the Haar Wavelet, and future research is required into whether this is the optimum filter for this task. Of particular interest is the question of whether higher order, non-symmetric or non-cubic filters might give even better results.
5. Conclusion
In this paper we have demonstrated a motion pre-processing algorithm which is both faster and more robust to noise than background modelling schemes. The algorithm works by detecting moving edges at various scales using spatio-temporal wavelet filters. The scaled down output of the algorithm also inherently aids object tracking, which requires a search for connected regions. The algorithm has been demonstrated to be considerably more sensitive for detection of motion even in cases where background modelling fails, where the contrast between background and target is low, or when the target is moving slowly, while at the same time maintaining a greater inherent robustness to noise. The algorithm can also track arbitrary objects without the need for a predefined target appearance model. Acknowledgment
This research has been funded by Research Councils UK under Basic Technology Grant GR/R87642/02. In addition, we acknowledge use of images in Figure 8 from EC Funded CAVIAR project/IST 2001 37540, found at URL: https://www.wendangku.net/doc/6a11013146.html,/rbf/CAVIAR/. References
[1] R. Cucchiara, C. Grana, M. Piccardi, A. Prati, “Detecting
moving objects, ghosts, and shadows in video streams”, IEEE Trans. Pattern Anal. Machine Intell., 25, no. 10, 2003, 1337–1342.
[2] T. Delbrück, “Silicon retina with correlation-based
velocity-tuned pixels”, IEEE Trans. Neural Networks, 4, no. 3, 1993, 529–541.
[3] R.R. Etienne-Cummings, J. van der Spiegel, P. Mueller, “A
focal plane visual motion measurement sensor”, IEEE Trans. Circuits and Systems I: Fundamental Theory and Applications, 44, no. 1, 1997, 55–66.
[4] M. Isard, J. MacCormick, “BraMBLe: A Bayesian
multiple-blob tracker”, Proc. Int. Conf. on Computer Vision (ICCV’01), Vol. 2, 2001, pp. 34–41.
[5] D. Makris, T. Ellis, “Finding paths in video sequences”,
Proc. British Machine Vision Conf. (BMVC’01), Sept.
2001, pp. 263–272.
[6] S. Nyiogo, E. Adelson, “Analyzing and Recognizing
Walking Figures in XYT”, MIT Media Lab, Vision Report
no. 223, 1994.
[7] E. Shechtman, M. Irani, “Space-time behavior based
correlation”, Proc. IEEE Computer Society Conf. on Computer Vision and Pattern Recognition (CVPR’05) - Vol. 1, 2005, pp. 405–412.
[8] C. Stauffer, W.E.L. Grimson. “Adaptive Background
Mixture Models for Real-Time Tracking”, IEEE Computer
Society Conference on Computer Vision and Pattern Recognition (CVPR'99) – Vol. 2, 1999, p. 2246.
[9] M. Sugrue, E.R. Davies, “Tracking in CCTV Video Using
Human Visual System Inspired Algorithm”, Visual Information Engineering 2005, University of Glasgow, 4–6
April 2005.
[10] T. Troscianko, A. Holmes, J. Stillman, M. Mirmehdi, D.
Wright, A. Wilson, “What happens next? The predictability
of natural behaviour viewed through CCTV cameras”, Perception, 33, no. 1, 2004, 87–101.
[11] Y. Wexler, E. Shechtman, and M. Irani, “Space-time video
completion”, Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR’04), June 2004, pp. 120–127.
古代晋灵公不君、齐晋鞌之战原文及译文
晋灵公不君（宣公二年） 原文： 晋灵公不君。厚敛以雕墙。从台上弹人，而观其辟丸也。宰夫胹熊蹯不熟，杀之，寘诸畚，使妇人载以过朝。赵盾、士季见其手，问其故而患之。将谏，士季曰：“谏而不入，则莫之继也。会请先，不入，则子继之。”三进及溜，而后视之，曰：“吾知所过矣，将改之。”稽首而对曰：“人谁无过？过而能改，善莫大焉。诗曰：‘靡不有初，鲜克有终。’夫如是，则能补过者鲜矣。君能有终，则社稷之固也，岂惟群臣赖之。又曰：‘衮职有阙，惟仲山甫补之。’能补过也。君能补过，衮不废矣。” 犹不改。宣子骤谏，公患之，使鉏麑贼之。晨往，寝门辟矣，盛服将朝。尚早，坐而假寐。麑退，叹而言曰：“不忘恭敬，民之主也。贼民之主，不忠；弃君之命，不信。有一于此，不如死也！”触槐而死。 秋九月，晋侯饮赵盾酒，伏甲将攻之。其右提弥明知之，趋登曰：“臣侍君宴，过三爵，非礼也。”遂扶以下。公嗾夫獒焉。明搏而杀之。盾曰：“弃人用犬，虽猛何为！”斗且出。提弥明死之。 初，宣子田于首山，舍于翳桑。见灵辄饿，问其病。曰：“不食三日矣！”食之，舍其半。问之，曰：“宦三年矣，未知母之存否。今近焉，请以遗之。”使尽之，而为之箪食与肉，寘诸橐以与之。既而与为公介，倒戟以御公徒，而免之。问何故，对曰：“翳桑之饿人也。”问其名居，不告而退。——遂自亡也。 乙丑，赵穿①攻灵公于桃园。宣子未出山而复。大史书曰：“赵盾弑其君。”以示于朝。宣子曰：“不然。”对曰：“子为正卿，亡不越竟，反不讨贼，非子而谁？”宣子曰：“呜呼!‘我之怀矣，自诒伊戚。’其我之谓矣。” 孔子曰：“董狐，古之良史也，书法不隐。赵宣子，古之良大夫也，为法受恶。惜也，越竞乃免。” 译文： 晋灵公不行君王之道。他向人民收取沉重的税赋以雕饰宫墙。他从高台上用弹弓弹人，然后观赏他们躲避弹丸的样子。他的厨子做熊掌，没有炖熟，晋灵公就把他杀了，把他的尸体装在草筐中，让宫女用车载着经过朝廷。赵盾和士季看到露出来的手臂，询问原由后感到很忧虑。他们准备向晋灵公进谏，士季说：“如果您去进谏而君王不听，那就没有人能够再接着进谏了。还请让我先来吧，不行的话，您再接着来。”士季往前走了三回，行了三回礼，一直到屋檐下，晋灵公才抬头看他。晋灵公说：“我知道我的过错了，我会改过的。”士季叩头回答道：“谁能没有过错呢？有过错而能改掉，这就是最大的善事了。《诗经》说：‘没有人向善没有一个开始的，但却很少有坚持到底的。’如果是这样，那么能弥补过失的人是很少的。您如能坚持向善，那么江山就稳固了，不只是大臣们有所依靠啊。
如何翻译古文
如何翻译古文 学习古代汉语，需要经常把古文译成现代汉语。因为古文今译的过程是加深理解和全面运用古汉语知识解决实际问题的过程，也是综合考察古代汉语水平的过程。学习古代汉语，应该重视古文翻译的训练。 古文翻译的要求一般归纳为信、达、雅三项。“信”是指译文要准确地反映原作的含义，避免曲解原文内容。“达”是指译文应该通顺、晓畅，符合现代汉语语法规范。“信”和“达”是紧密相关的。脱离了“信”而求“达”，不能称为翻译；只求“信”而不顾“达”，也不是好的译文。因此“信”和“达”是文言文翻译的基本要求。“雅”是指译文不仅准确、通顺，而且生动、优美，能再现原作的风格神韵。这是很高的要求，在目前学习阶段，我们只要能做到“信”和“达”就可以了。 做好古文翻译，重要的问题是准确地理解古文，这是翻译的基础。但翻译方法也很重要。这里主要谈谈翻译方法方面的问题。 一、直译和意译 直译和意译是古文今译的两大类型，也是两种不同的今译方法。 1．关于直译。所谓直译，是指紧扣原文，按原文的字词和句子进行对等翻译的今译方法。它要求忠实于原文，一丝不苟，确切表达原意，保持原文的本来面貌。例如： 原文：樊迟请学稼，子曰：“吾不如老农。”请学为圃。子曰：“吾不如老圃。”（《论语?子路》） 译文：樊迟请求学种庄稼。孔子道：“我不如老农民。”又请求学种菜蔬。孔子道：“我不如老菜农。”（杨伯峻《论语译注》） 原文：齐宣王问曰：“汤放桀，武王伐纣，有诸？”（《孟子?梁惠王下》） 译文：齐宣王问道：“商汤流放夏桀，武王讨伐殷纣，真有这回事吗？（杨伯峻《孟子译注》） 上面两段译文紧扣原文，字词落实，句法结构基本上与原文对等，属于直译。 但对直译又不能作简单化理解。由于古今汉语在文字、词汇、语法等方面的差异，今译时对原文作一些适当的调整，是必要的，并不破坏直译。例如： 原文：逐之，三周华不注。（《齐晋鞌之战》） 译文：〔晋军〕追赶齐军，围着华不注山绕了三圈。
齐晋鞌之战原文和译文
鞌之战选自《左传》又名《鞍之战》原文：楚癸酉，师陈于鞌(1)。邴夏御侯，逢丑父为右②。晋解张御克，郑丘缓为右(3)。侯日：“余姑翦灭此而朝食(4)”。不介马而驰之⑤。克伤于矢，流血及屦2 未尽∧6)，曰：“余病矣(7)！”张侯曰：“自始合(8)，而矢贯余手及肘(9)，余折以御，左轮朱殷(10)，岂敢言病吾子忍之！”缓曰：“自始合，苟有险，余必下推车，子岂＿识之(11)然子病矣！”张侯曰：“师之耳目，在吾旗鼓，进退从之。此车一人殿之(12)，可以集事(13)，若之何其以病败君之大事也擐甲执兵(14)，固即死也(15)；病未及死，吾子勉之(16)！”左并辔(17) ，右援拐鼓(18)。马逸不能止(19)，师从之，师败绩。逐之，三周华不注(20) 韩厥梦子舆谓己曰：“旦辟左右！”故中御而从齐侯。邴夏曰：“射其御者，君子也。”公曰：“谓之君子而射之，非礼也。”射其左，越于车下；射其右，毙于车中。綦毋张丧车，从韩厥，曰：“请寓乘。”从左右，皆肘之，使立于后。韩厥俛，定其右。逢丑父与公易位。将及华泉，骖絓于木而止。丑父寝于轏中，蛇出于其下，以肱击之，伤而匿之，故不能推车而及。韩厥执絷马前，再拜稽首，奉觞加璧以进，曰：“寡君使群臣为鲁、卫请，曰：‘无令舆师陷入君地。’下臣不幸，属当戎行，无所逃隐。且惧奔辟而忝两君，臣辱戎士，敢告不敏，摄官承乏。” 丑父使公下，如华泉取饮。郑周父御佐车，宛茷为右，载齐侯以免。韩厥献丑父，郤献子将戮之。呼曰：“自今无有代其君任患者，有一于此，将为戮乎”郤子曰：“人不难以死免其君，我戮之不祥。赦之，以劝事君者。”乃免之。译文1：在癸酉这天，双方的军队在鞌这个地方摆开了阵势。齐国一方是邴夏为齐侯赶车，逢丑父当车右。晋军一方是解张为主帅郤克赶车，郑丘缓当车右。齐侯说:“我姑且消灭了这些人再吃早饭。”不给马披甲就冲向了晋军。郤克被箭射伤，血流到了鞋上，但是仍不停止擂鼓继续指挥战斗。他说:“我受重伤了。”解张说:“从一开始接战，一只箭就射穿了我的手和肘，左边的车轮都被我的血染成了黑红色，我哪敢说受伤您忍着点吧！”郑丘缓说:“从一开始接战，如果遇到道路不平的地方，我必定（冒着生命危险）下去推车，您难道了解这些吗不过，您真是受重伤了。”daier 解张说:“军队的耳朵和眼睛，都集中在我们的战旗和鼓声，前进后退都要听从它。这辆车上还有一个人镇守住它，战事就可以成功。为什么为了伤痛而败坏国君的大事呢身披盔甲，手执武器，本来就是去走向死亡，伤痛还没到死的地步，您还是尽力而为吧。”一边说，一边用左手把右手的缰绳攥在一起，用空出的右手抓过郤克手中的鼓棰就擂起鼓来。（由于一手控马，）马飞快奔跑而不能停止，晋军队伍跟着指挥车冲上去，把齐军打得打败。晋军随即追赶齐军，三次围绕着华不注山奔跑。韩厥梦见他去世的父亲对他说:“明天早晨作战时要避开战车左边和右边的位置。”因此韩厥就站在中间担任赶车的来追赶齐侯的战车。邴夏说:“射那个赶车的，他是个君子。”齐侯说: “称他为君子却又去射他，这不合于礼。”daier 于是射车左，车左中箭掉下了车。又射右边的，车右也中箭倒在了车里。（晋军的）将军綦毋张损坏了自己的战车，跟在韩厥的车后说: “请允许我搭乗你的战车。”他上车后，无论是站在车的左边，还是站在车的右边，韩厥都用肘推他，让他站在自己身后——战车的中间。韩厥又低下头安定了一下受伤倒在车中的那位自己的车右。于是逢丑父和齐侯（乘韩厥低头之机）互相调换了位置。将要到达华泉时，齐侯战车的骖马被树木绊住而不能继续逃跑而停了下来。（头天晚上）逢丑父睡在栈车里，有一条蛇从他身子底下爬出来，他用小臂去打蛇，小臂受伤，但他（为了能当车右）隐瞒了这件事。由于这样，他不能用臂推车前进，因而被韩厥追上了。韩厥拿着拴马绳走到齐侯的马前，两次下拜并行稽首礼，捧着一杯酒并加上一块玉璧给齐侯送上去，
微生物实验室常用仪器配置
微生物实验室常用仪器配置 1 超净工作台（已有）微生物的培养都是在特定培养基中进行无菌培养，那么无菌培养必然需要超净工作台提供一个无菌的工作环境 2、培养箱（已有）培养箱有多种类型，它的作用在于为微生物的生长提供一个适宜的环境。生化培养箱只能控制温度，可作为一般细菌的平板培养;霉菌培养箱可以控制温度和湿度，可作为霉菌的培养;CO2培养箱适用于厌氧微生物的培养。 3、天平（已有待完善）天平用于精确称量各类试剂。实验室常用的是电子天平，电子天平按照精度不同有不同的级别。 4、微生物均质器（无）用于从固体样品中提取细菌。用微生物均质器制备微生物检测样本具有样品无污染、无损伤、不升温、不需要灭菌处理，不需洗刷器皿等特点，是微生物实验中使用较为方便的仪器 5、菌落计数器（无）菌落计数仪可协助操作者计数菌落数量。通过放大，拍照，计数等方式准确的获取菌落的数量。有些高性能的菌落计数器还可连接电脑完成自动计数的操作。 微波炉/ 电炉用于溶液的快速加热，微生物固体培养基的加热溶化。 7、高压灭菌锅（已有坏）微生物学所用到的大部分实验物品、试剂、培养基都应严格消毒灭菌。灭菌锅也有不同大小型号，有些是手动的，有些是全自动的。用户需要根据自己的需要选购。 8、移液器（无）液体量器用于精密量取各类液体。常见的液体量器有量筒、移液管、微量取液器、刻度试管、烧杯。 9、低温冰箱（已有）冰箱是实验室保存试剂和样品必不可少的仪器。微生物学实验中用到的试剂有些要求是 4 度保存，有些要求是负20 度保存，实验人员一定要看清试剂的保存条件，放置在恰当的温度下保存。 11、摇床（已有）摇床是实验室常用的一种仪器，在微生物实验操作过程中，液体培养基培养细菌时需要在特定温度下振荡使用。 12、纯水装置 13、生物显微镜（无）由于微生物体积较小，所以在观察时需要借助生物显微镜。生物显微镜用于微生物和微小物品结构，形态等的观察。 16、恒温干燥箱 17、恒温水浴锅（已有------ 其他实验室） 水浴锅是一种控温装置，水浴控温对于样品来说比较快速且接触充分。有些微生物反应需要在37度，42度，56度下水浴进行，所以恒温水浴锅可以提供需要的温度。 （18）酸度计（无） 用于配置试剂时精确测量PH 值，从而保证配置的溶液的精确性。有时也需要利用pH 计测定样品溶液的酸碱度。 （19）离心机（无）用于收集微生物菌体以及其他沉淀物。离心机有冷冻和常温之分。 有些样品由于在常温下不太稳定，需要低温环境，要视样品的种类而定。 20 接种仪器：试管（有）、接种针、牛角勺（无）、药勺、接种环、接种钩，解剖刀、培养皿（无）、镊子、酒精灯、接种锄、玻璃刮刀（无）
《鞌之战》阅读答案(附翻译)原文及翻译
《鞌之战》阅读答案（附翻译）原文及翻 译 鞌之战[1] 选自《左传成公二年（即公元前589年）》 【原文】 癸酉，师陈于鞌[2]。邴夏御齐侯[3]，逢丑父为右[4]。晋解张御郤克，郑丘缓为右[5]。齐侯曰：余姑翦灭此而朝食[6]。不介马而驰之[7]。郤克伤于矢，流血及屦，未绝鼓音[8]，曰：余病[9]矣！张侯[10]曰：自始合，而矢贯余手及肘[11]，余折以御，左轮朱殷[12]，岂敢言病。吾子[13]忍之！缓曰：自始合，苟有险[14]，余必下推车，子岂识之[15]？然子病矣！张侯曰：师之耳目，在吾旗鼓，进退从之[16]。此车一人殿之[17]，可以集事[18]，若之何其以病败君之大事也[19]？擐甲执兵，固即死也[20]。病未及死，吾子勉之[21]！左并辔[22]，右援枹而鼓[23]，马逸不能止[24]，师从之。齐师败绩[25]。逐之，三周华不注[26]。 【注释】 [1]鞌之战：春秋时期的著名战役之一。战争的实质是齐、晋争霸。由于齐侯骄傲轻敌，而晋军同仇敌忾、士气旺盛，战役以齐败晋胜而告终。鞌：通鞍，齐国地名，在今山东济南西北。 [2]癸酉：成公二年的六月十七日。师，指齐晋两国军队。陈，
列阵，摆开阵势。 [3]邴夏：齐国大夫。御，动词，驾车。御齐侯，给齐侯驾车。齐侯，齐国国君，指齐顷公。 [4]逢丑父：齐国大夫。右：车右。 [5]解张、郑丘缓：都是晋臣，郑丘是复姓。郤（x ）克，晋国大夫，是这次战争中晋军的主帅。又称郤献子、郤子等。 [6]姑：副词，姑且。翦灭：消灭，灭掉。朝食：早饭。这里是吃早饭的意思。这句话是成语灭此朝食的出处。 [7]不介马：不给马披甲。介：甲。这里用作动词，披甲。驰之：驱马追击敌人。之：代词，指晋军。 [8] 未绝鼓音：鼓声不断。古代车战，主帅居中，亲掌旗鼓，指挥军队。兵以鼓进，击鼓是进军的号令。 [9] 病：负伤。 [10]张侯，即解张。张是字，侯是名，人名、字连用，先字后名。 [11]合：交战。贯：穿。肘：胳膊。 [12]朱：大红色。殷：深红色、黑红色。 [13]吾子：您，尊敬。比说子更亲切。 [14]苟：连词，表示假设。险：险阻，指难走的路。 [15]识：知道。之，代词，代苟有险，余必下推车这件事，可不译。 [16]师之耳目：军队的耳、目（指注意力）。在吾旗鼓：在我们
设备检测实验室设备配置建议最新
第十八章无线电设备检测实验室设备配置建议 一、概述： 无线电设备检测工作是无线电管理的一个重要方面，随着科学技术水平的飞速发展，无线电通信应用愈来愈广泛，各种无线电通信设备迅速增加，新技术、新体制、新设备不断涌现，越来越多的无线电通信技术在我国得以应用。为了使有限的频谱资源能够科学地、有效地开发和利用，防止无线电设备本身产品质量不合格产生的各种有害干扰，从源头上减少无线电干扰信号，维护空中电波秩序，确保各种无线电设备正常进行，必须加强对各类无线电设备的管理，应积极开展无线电设备检测工作。 现阶段在我国应用的通信技术体制非常广泛，目前 TDMA(GSM900/DCS1800/GPRS)、CDMA(cdmaOne)、PHS(小灵通)为代表的通信体制正在被中国移动、中国联通、中国电信、中国网通广泛的应用。同时，第三代（3G）移动通信系统(cdma2000、W-CDMA、TD-S-CDMA、LAS-CDMA 等)也正在紧锣密鼓的研究开发之中，在不远的将来可能得到应用。在当前的数字时代，除电信运营企业全部采用数字调制技术实现运营网络的数据、语音、图象等的无线传输外，其它应用场合也越来越多地应用数字调制技术，如DBS直接广播卫星通信、数字集群通信、Bluetooh蓝牙数据传输、WirelessLAN无线区域通信等。同时，各种新的系统如LMDS，WLAN，BLUETOOTH等不断投入使用。必将对无线电设备检测工作提出了更高的要求。我们认为，作为无线电频谱的管理部门，监管的重点首先是无线发射机系统的射频指标，如:发射总功率、频率准确度、占用带宽、杂散发射
等。在此基础上，可适当进行其他指标的测试，辅助运营商进行系统检测。其主要测试项目应包括： 1.频率准确度 2.最大输出功率 3.发射机互调 4.杂散发射 5.占用带宽 OBW、邻道功率ACP 6.相位误差 另外，针对GSM系统，可增加调制及开关的频谱，功率时间曲线的测试；针对CDMA系统和扩频系统，可增加矢量幅度误差（EVM），码域功率和幅度统计特性（CCDF）等测试项目。 为了确保检测工作的准确性、权威性，必须建立一个合格的检测实验室。它必须具备有效的质量管理体系、完全满足被测设备技术指标的测量仪器、经过培训的专业技术人员。以下重点讨论检测实验室需配备的仪器。 二、检测实验室仪器仪表的配置应原则： 1．仪器功能强：应提供符合国家标准规范所要求的主要项目和指标。 2．测量基准高：测量结果应具备权威性，以便对网络设备和用户终端进行认证、验收、日常抽查、事故判别和质量检测。 3．通用性能好：应尽量基于单台设备平台对各种体制网络设备（TDMA、CDMA、PHS、CDMA2000、W-CDMA等）进行测量。 4．方便携带性：可以方便地在野外现场搭建测量系统。运输、搬移方便，抗震动和适应恶劣环境的能力强。
《鞌之战》阅读答案附翻译
《鞌之战》阅读答案（附翻译） 《鞌之战》阅读答案（附翻译） 鞌之战[1] 选自《左传·成公二年（即公元前589年）》 【原文】 癸酉，师陈于鞌[2]。邴夏御齐侯[3]，逢丑父为右[4]。晋解张御郤克，郑丘缓为右[5]。齐侯曰：“余姑 翦灭此而朝食[6]。”不介马而驰之[7]。郤克伤于矢， 流血及屦，未绝鼓音[8]，曰：“余病[9]矣！”张侯[10]曰：“自始合，而矢贯余手及肘[11]，余折以御，左轮 朱殷[12]，岂敢言病。吾子[13]忍之！”缓曰：“自始合，苟有险[14]，余必下推车，子岂识之[15]？——然 子病矣！”张侯曰：“师之耳目，在吾旗鼓，进退从之[16]。此车一人殿之[17]，可以集事[18]，若之何其以 病败君之大事也[19]？擐甲执兵，固即死也[20]。病未 及死，吾子勉之[21]！”左并辔[22]，右援枹而鼓[23]，马逸不能止[24]，师从之。齐师败绩[25]。逐之，三周 华不注[26]。 【注释】 [1]鞌之战：春秋时期的著名战役之一。战争的实质是齐、晋争霸。由于齐侯骄傲轻敌，而晋军同仇敌忾、 士气旺盛，战役以齐败晋胜而告终。鞌：通“鞍”，齐
国地名，在今山东济南西北。 [2]癸酉：成公二年的六月十七日。师，指齐晋两国军队。陈，列阵，摆开阵势。 [3]邴夏：齐国大夫。御，动词，驾车。御齐侯，给齐侯驾车。齐侯，齐国国君，指齐顷公。 [4]逢丑父：齐国大夫。右：车右。 [5]解张、郑丘缓：都是晋臣，“郑丘”是复姓。郤（xì）克，晋国大夫，是这次战争中晋军的主帅。又称郤献子、郤子等。 [6]姑：副词，姑且。翦灭：消灭，灭掉。朝食：早饭。这里是“吃早饭”的意思。这句话是成语“灭此朝食”的出处。 [7]不介马：不给马披甲。介：甲。这里用作动词，披甲。驰之：驱马追击敌人。之：代词，指晋军。 [8]未绝鼓音：鼓声不断。古代车战，主帅居中，亲掌旗鼓，指挥军队。“兵以鼓进”，击鼓是进军的号令。 [9]病：负伤。 [10]张侯，即解张。“张”是字，“侯”是名，人名、字连用，先字后名。 [11]合：交战。贯：穿。肘：胳膊。 [12]朱：大红色。殷：深红色、黑红色。 [13]吾子：您，尊敬。比说“子”更亲切。
微生物实验室常用仪器配置
微生物实验室常用仪器配置 微生物学实验室是生物学领域的一个基本实验室，对于一个完备的微生物学实验室，我们需要配置哪些仪器呢?环凯为您的微生物学实验仪器配置提供如下参考。 1、超净工作台 微生物的培养都是在特定培养基中进行无菌培养，那么无菌培养必然需要超净工作台提供一个无菌的工作环境。 2、培养箱 培养箱有多种类型，它的作用在于为微生物的生长提供一个适宜的环境。生化培养箱只能控制温度，可作为一般细菌的平板培养;霉菌培养箱可以控制温度和湿度，可作为霉菌的培养;CO2培养箱适用于厌氧微生物的培养。 3、天平 天平用于精确称量各类试剂。实验室常用的是电子天平，电子天平按照精度不同有不同的级别。 4、微生物均质器
用于从固体样品中提取细菌。用微生物均质器制备微生物检测样本具有样品无污染、无损伤、不升温、不需要灭菌处理，不需洗刷器皿等特点，是微生物实验中使用较为方便的仪器。 5、菌落计数器 菌落计数仪可协助操作者计数菌落数量。通过放大，拍照，计数等方式准确的获取菌落的数量。有些高性能的菌落计数器还可连接电脑完成自动计数的操作。 6、微波炉/电炉 用于溶液的快速加热，微生物固体培养基的加热溶化。 7、高压灭菌锅 微生物学所用到的大部分实验物品、试剂、培养基都应严格消毒灭菌。灭菌锅也有不同大小型号，有些是手动的，有些是全自动的。用户需要根据自己的需要选购。 8、移液器 液体量器用于精密量取各类液体。常见的液体量器有量筒、移液管、微量取液器、刻度试管、烧杯。 9、低温冰箱 冰箱是实验室保存试剂和样品必不可少的仪器。微生物学实验中用到的试剂有些要求是4度保存，有些要求是负20度保存，实验人员一定要看清试剂的保存条件，放置在恰当的温度下保存。 10、生物安全柜 微生物实验中涉及的试剂和样品微生物有些是有毒的，对于操作人员来说伤害较大。为了防止有害悬浮微粒、气溶胶的扩散，可以利用生物安全柜对操作人员、样品及样品间交叉感染和环境提供安全保护。 11、摇床 摇床是实验室常用的一种仪器，在微生物实验操作过程中，液体培养基培养细菌时需要在特定温度下振荡使用。 12、纯水装置
左传《齐晋鞌之战》原文+翻译+注释
左传《齐晋鞌之战》原文+翻译+注释 楚癸酉，师陈于鞌(1)。邴夏御侯，逢丑父为右②。晋解张御克，郑丘缓 为右(3)。侯日：“余姑翦灭此而朝食(4)”。不介马而驰之⑤。克伤于矢， 流血及屦2未尽∧?6)，曰：“余病矣(7)！”张侯曰：“自始合(8)，而矢贯余手 及肘(9)，余折以御，左轮朱殷(10)，岂敢言病？吾子忍之！”缓曰：“自始合，苟有险，余必下推车，子岂＿识之(11)？然子病矣！”张侯曰：“师之耳目，在 吾旗鼓，进退从之。此车一人殿之(12)，可以集事(13)，若之何其以病败君之大事也？擐甲执兵(14)，固即死也(15)；病未及死，吾子勉之(16)！”左并辔(17) ，右援拐?鼓(18)。马逸不能止(19)，师从之，师败绩。逐之，三周华不注(20) 韩厥梦子舆谓己曰：“旦辟左右！”故中御而从齐侯。邴夏曰：“射其御者，君子也。”公曰：“谓之君子而射之，非礼也。”射其左，越于车下；射其右，毙于车中。綦毋张丧车，从韩厥，曰：“请寓乘。”从左右，皆肘之，使立于后。韩厥俛，定其右。逢丑父与公易位。将及华泉，骖絓于木而止。丑父寝于轏中，蛇出于其下，以肱击之，伤而匿之，故不能推车而及。韩厥执絷马前，再拜稽首，奉觞加璧以进，曰：“寡君使群臣为鲁、卫请，曰：‘无令舆师陷入君地。’下臣不幸，属当戎行，无所逃隐。且惧奔辟而忝两君，臣辱戎士，敢告不敏，摄官承乏。”丑父使公下，如华泉取饮。郑周父御佐车，宛茷为右，载齐侯以免。韩厥献丑父，郤献子将戮之。呼曰：“自今无有代其君任患者，有一于此，将为戮乎？”郤子曰：“人不难以死免其君，我戮之不祥。赦之，以劝事君者。”乃免之。 在癸酉这天，双方的军队在鞌这个地方摆开了阵势。齐国一方是邴夏为齐侯赶车，逢丑父当车右。晋军一方是解张为主帅郤克赶车，郑丘缓当车右。齐侯说:“我姑且消灭了这些人再吃早饭。”不给马披甲就冲向了晋军。郤克被箭射伤，血流到了鞋上，但是仍不停止擂鼓继续指挥战斗。他说:“我受重伤了。”解张说:“从一开始接战，一只箭就射穿了我的手和肘，左边的车轮都被我的血染成了黑红色，我哪敢说受伤？您忍着点吧！”郑丘缓说:“从一开始接战，如果遇到道路不平的地方，我必定（冒着生命危险）下去推车，您难道了解这些吗？不过，您真是受重伤了。”daier解张说:“军队的耳朵和眼睛，都集中在我们的战旗和鼓声，前进后退都要听从它。这辆车上还有一个人镇守住它，战事就可以成功。为什么为了伤痛而败坏国君的大事呢？身披盔甲，手执武器，本来就是去走向死亡，伤痛还没到死的地步，您还是尽力而为吧。”一边说，一边用左手把右手的缰绳攥在一起，用空出的右手抓过郤克手中的鼓棰就擂起鼓来。（由于一手控马，）马飞快奔跑而不能停止，晋军队伍跟着指挥车冲上去，把齐军打得打败。晋军随即追赶齐军，三次围绕着华不注山奔跑。
分子生物学实验室需要的仪器配置
分子生物学实验室需要的仪器配置 (1)培养箱在分子生物学试验中，有很多反应都是在特定温度下进行的，这时就需要一个控温的装置。例如：用于细菌的平板培养，我们通常设定为37℃于培养箱倒置培养；其他分子生物学实验如酶切等需要25℃，30℃，37℃等条件。 (2)冰箱冰箱是实验室保存试剂和样品必不可少的仪器。分子生物学实验中用到的试剂有些要求是4度保存，有些要求是负20度保存，实验人员一定要看清试剂的保存条件，放置在恰当的温度下保存。具体来说，不同温度下保存的物品如下： a. 4℃适合储存某些溶液、试剂、药品等。 b.-20℃适用于某些试剂、药品、酶、血清、配好的抗生素和DNA、蛋白质样品等。 c.-80℃适合某些长期低温保存的样品、大肠杆菌菌种、纯化的样品、特殊的低温处理消化液，感受态等的保存。 d.0-10℃的层析冷柜适合低温条件下的电泳、层析、透析等实验。 (3)摇床摇床是实验室常用仪器，一般有常温型和低温型两种。对于分子生物学实验室，如果能配置低温型摇床，就可以适应不同的实验需求。例如：用于大肠杆菌，酵母菌等生物工程菌种的振荡培养及蛋白的诱导表达，培养通常为28度和37度，诱导表达需要20-37度；在感受态的制备过程中，需要有18度的温度控制；用于蛋白凝胶的染色脱色时振荡，常温使用；用于大肠杆菌常规转化时振荡复苏，常为37度。对于控制温度低于室温时，我们需要低温型摇床来控温。 (4)水浴锅水浴锅也是一种控温装置，水浴控温对于样品来说比较快速且接触充分。例如，用于42度的大肠杆菌转化时的热激反应；用于DNA杂交过程中水浴控温。 (5)烘箱烘箱是用于灭菌和洗涤后的物品烘干。烘箱有不同的控温范围，用户可以根据实验需求进行选择。例如，有些塑料用具只能在42-45℃的烤箱中进行烘干；用于RNA方面的实验用具，需要在250℃烤箱中烘干。 (6)纯水装置纯水装置包括蒸馏水器和纯水机。蒸馏水器的价格便宜，但在造水过程中需要有人值守；纯水机价格高些，但是使用方便，可以储存一定量的纯水。纯水使用也有不同的级别，一般实验用水需要纯水，用于PCR、DNA测序、酶反应均需要超纯水。 (7)灭菌锅分子生物学所用到的大部分实验用具都应严格消毒灭菌。包括实验物品、试剂、培养基等。灭菌锅也有不同大小型号，有些是手动的，有些是全自动的。用户需要根据自己的需要选购。 (8)天平天平用于精确称量各类试剂。实验室常用的是电子天平，电子天平按照精度不同有不同的级别。
各行业实验室仪器基本配置
各行业实验室建设所需仪器基本配置 各行业实验室仪器基本配置 供水公司 仪器名称具体应用 BOD测定仪测量水中生物需氧量 双道原子荧光光度计可检测多种元素 溶解氧测定仪测溶解氧 数字浊度计测量液体浊度 原子吸收分光光度计根据被测元素的基态原子对特征辐射的吸收程度进行定量分析实验室专用超纯水机清洗玻璃器皿、配置标准溶液、理化分析等，制备纯水和超纯水电导率仪测电解质溶液电导率值 气相色谱仪定性、定量分析 多功能红外测油仪测量水中油含量 二氧化碳分析仪分析二氧化碳含量 放射性测量仪测量水中放射性核素 分光光度计定量分析 傅里叶红外变换光谱仪定性分析 光电式浑浊度仪测水的浑浊度 总有机碳测定仪对水溶液中总有机碳进行定量测定 离子色谱仪适用于亲水性阴、阳离子的分离 酸度计测ＰＨ值 显微镜观察微小物质 火焰光度计测定元素含量 激光粉尘仪测空气中粉尘含量 露点仪测露点 冷原子荧光测汞仪测汞含量 微量氧分析仪测微量的氧气浓度 油份浓度分析仪测定水中油的含量 水气厂 仪器名称具体应用 离子色谱仪适用于亲水性阴、阳离子的分离 液相色谱仪定性、定量分析 实验室专用超纯水机清洗玻璃器皿、配置标准溶液、理化分析等，制备纯水和超纯水ＴＯＰＣ测定仪测量源水、饮水和超纯水的总有机碳含量 火焰光度计分析二氧化碳含量 浊度计测定元素含量 露点仪测量液体浊度 微量氧分析仪测露点 原子吸收分光光度计测定水中微量氧
油分浓度分析仪根据被测元素的基态原子对特征辐射的吸收程度进行定量分析 常量氧分析仪测定水中油的含量 多功能红外测油仪测定水中氧 傅里叶红外变换光谱仪定性分析 溶解氧分析仪测溶解氧 农产品质检站 仪器名称具体应用 酸度计测pＨ值 电导率仪测电解质溶液电导率值 液相色谱仪定性、定量分析 实验室专用超纯水机清洗玻璃器皿、配置标准溶液、理化分析等，制备纯水和超纯水 气相色谱仪定性、定量分析 自动电位滴定仪酸碱滴定、氧化还原滴定、沉淀滴定、络合滴定 紫外—可见分光光度计测量物质对不同波长单色辐射的吸收程度，定量分析 可见分光光度计测量物质对不同波长单色辐射的吸收程度，定量分析 原子吸收分光光度计根据被测元素的基态原子特征辐射的吸收程度进行定量分析 红外分光光度计根据物质在红外光区的吸收光谱特征和朗伯比尔定律对物质进行定性定量 分析 卡尔费休水份仪测定含水的仪器 傅立叶变换红外光谱仪定性、定量的分析 色差计定性分析 离子色谱仪检测农家产品中的微量水份 微量水份测定仪用于易形成氢化物元素、易形成气态组分元素和易还原成原子蒸汽元素的测 定 荧光分光光度计测物质旋光度，分析物质的浓度、纯度、含糖量 旋光仪（目视、自动）根据酶与底物能产生显色反应，对底物进行研究定性及定量分析 酶标分析仪移取微量溶液 微量移液器农产品中物质的定性定量分析 气质联用仪测折射率 阿贝尔折射仪测面粉、淀粉等粉剂的白度值 白度计氨基酸含量 氨基酸自动分析仪通过用标准色比较测颜色 比较测色仪用未知浓度样品与已知浓度标物比较方法进行定量分析 比色计测定粮食中含水量 电脑粮食水分仪测蛋白质中氮的含量 凯氏定氮仪测谷物中含水量 谷物水份仪测黄曲霉素含量 黄曲霉素测定仪测甲醛、氨含量 甲醛氨测定仪测定空气中甲醛气体的含量 甲醛测定仪测定谷物、面粉及其它含有淀粉的产品中淀粉酶活性 降落值测定仪测混浊度 浊度仪测含糖量、糖度
《鞌之战》阅读答案(附翻译)
鞌之战[1]选自《左传·成公二年（即公元前589年）》【原文】癸酉，师陈于鞌[2]。邴夏御齐侯[3]，逢丑父为右[4]。晋解张御郤克，郑丘缓为右[5]。齐侯曰：“余姑翦灭此而朝食[6]。”不介马而驰之[7]。郤克伤于矢，流血及屦，未绝鼓音[8]，曰：“余病[9]矣！”张侯[10]曰：“自始合，而矢贯余手及肘[11]，余折以御，左轮朱殷[12]，岂敢言病。吾子[13]忍之！”缓曰：“自始合，苟有险[14]，余必下推车，子岂识之[15]？——然子病矣！”张侯曰：“师之耳目，在吾旗鼓，进退从之[16]。此车一人殿之[17]，可以集事[18]，若之何其以病败君之大事也[19]？擐甲执兵，固即死也[20]。病未及死，吾子勉之[21]！”左并辔[22]，右援枹而鼓[23]，马逸不能止[24]，师从之。齐师败绩[25]。逐之，三周华不注[26]。【注释】 [1]鞌之战：春秋时期的著名战役之一。战争的实质是齐、晋争霸。由于齐侯骄傲轻敌，而晋军同仇敌忾、士气旺盛，战役以齐败晋胜而告终。鞌：通“鞍”，齐国地名，在今山东济南西北。 [2]癸酉：成公二年的六月十七日。师，指齐晋两国军队。陈，列阵，摆开阵势。 [3]邴夏：齐国大夫。御，动词，驾车。御齐侯，给齐侯驾车。齐侯，齐国国君，指齐顷公。 [4]逢丑父：齐国大夫。右：车右。 [5]解张、郑丘缓：都是晋臣，“郑丘”是复姓。郤（xì）克，晋国大夫，是这次战争中晋军的主帅。又称郤献子、郤子等。 [6]姑：副词，姑且。翦灭：消灭，灭掉。朝食：早饭。这里是“吃早饭”的意思。这句话是成语“灭此朝食”的出处。 [7]不介马：不给马披甲。介：甲。这里用作动词，披甲。驰之：驱马追击敌人。之：代词，指晋军。 [8] 未绝鼓音：鼓声不断。古代车战，主帅居中，亲掌旗鼓，指挥军队。“兵以鼓进”，击鼓是进军的号令。 [9] 病：负伤。 [10]张侯，即解张。“张”是字，“侯”是名，人名、字连用，先字后名。 [11]合：交战。贯：穿。肘：胳膊。 [12]朱：大红色。殷：深红色、黑红色。 [13]吾子：您，尊敬。比说“子”更亲切。 [14]苟：连词，表示假设。险：险阻，指难走的路。 [15]识：知道。之，代词，代“苟有险，余必下推车”这件事，可不译。 [16]师之耳目：军队的耳、目（指注意力）。在吾旗鼓：在我们的旗子和鼓声上。进退从之：前进、后退都听从它们。 [17]殿之：镇守它。殿：镇守。 [18]可以集事：可以（之）集事，可以靠它（主帅的车）成事。集事：成事，指战事成功。 [19]若之何：固定格式，一般相当于“对……怎么办”“怎么办”。这里是和语助词“其”配合，放在谓语动词前加强反问，相当于“怎么”“怎么能”。以，介词，因为。败，坏，毁坏。君，国君。大事，感情。古代国家大事有两件：祭祀与战争。这里指战争。 [20]擐：穿上。执兵，拿起武器。 [21]勉，努力。 [22]并，动词，合并。辔（pèi）：马缰绳。古代一般是四匹马拉一车，共八条马缰绳，两边的两条系在车上，六条在御者手中，御者双手执之。“左并辔”是说解张把马缰绳全合并到左手里握着。 [23]援:拿过来。枹（fú）:击鼓槌。鼓:动词，敲鼓。 [24]逸:奔跑，狂奔。 [25] 败绩：大败。 [26] 周：环绕。华不注：山名，在今山东济南东北。【译文】六月十七日，齐晋两军在鞌地摆开阵势。邴夏为齐侯驾车，逢丑父担任车右做齐侯的护卫。晋军解张替郤克驾车，郑丘缓做了郤克的护卫。齐侯说：“我姑且消灭了晋军再吃早饭！”齐军没有给马披甲就驱车进击晋军。郤克被箭射伤，血一直流到鞋上，但他一直没有停止击鼓进。并说：“我受重伤了！”解张说：“从开始交战，箭就射穿了我的手和胳膊肘，我折断箭杆继续驾车，左边的车轮被血染得深红色，哪里敢说受了重伤？您还是忍住吧。”郑丘缓说：“从开始交战，只要遇到险峻难走的路，我必定要下去推车，您哪里知道这种情况呢？——不过您确实受重伤了！”解张说：“我们的旗帜和战鼓是军队的耳目，或进或退都听从旗鼓指挥。这辆战车只要一人镇守，就可以凭它成事。怎么能因为受伤而败坏国君的大事呢？穿上铠甲，拿起武器，本来就抱定了必死的决心。您虽然受了重伤还没有到死的地步，您就尽最大的努力啊！”于是左手把马缰绳全部握在一起，右手取过鼓槌来击鼓。战马狂奔不止，晋军跟着主帅的车前进。齐军大败，晋军追击齐军，绕着华不注山追了三圈。
食品实验室设备配置清单(原创)
实验室常用设备 1 常规理化分析实验室： 名 称数 量 电热恒温水浴锅2 旋转蒸发器2 真空循环水泵1 电热恒温干燥箱1 真空恒温干燥箱1 超声波提取/清洗器1 索式提取器2 挥发油提取器（轻 2 油） 电子数字天平1 PH酸度计1 磁力搅拌器1 超纯水器1 玻璃仪器及其他常用仪器： 玻璃干燥器、布氏漏斗、过滤瓶、培养皿、加料漏斗、烧杯、量筒、 具塞三角瓶、三角瓶、具塞抽滤瓶、茄形瓶、单口圆底烧瓶、三口烧瓶、表面皿、储液球、塑料洗瓶、普通干燥器、玻璃棒、四氟吸棒导气管、烧瓶托、抽滤套、温度计、注射器、红皮头、双连球、酒精灯、石棉网、镊子、四氟搅拌棒、刮刀、托盘、加热电圈、冷凝管夹、德式十字架、中铁台、铁圈、电热套调温控温、变压器、磁子、标签纸、自封袋、优质真空管、玻璃刀、毛细管、座式酒精灯、手套、干燥塔、球形冷凝管、直形冷凝管、层析柱、一球干燥管、斜行干燥管、蒸馏头75度、搅拌器套管、90度抽气接头、真空接收管、四氟搅拌器塞头、转接头、防溅球、布氏漏斗、三角漏斗、分液漏斗、恒压漏斗、具塞二通活塞、分水器、高效板、刻度试管、玻璃研钵、滤纸、脱脂棉、真空硅酯、螺旋夹、烧瓶夹、十字夹、甲基硅油、牛角匙、点样管、温度计、PH试纸、称量纸、试管刷、试管夹、生胶带、电吹风、封口膜、酒精喷灯、滴管、吸管、样品瓶、刮勺、搪瓷盘、离心管、分馏头、梨
形漏斗、展开槽/染色缸、西林瓶（青霉素瓶）及塞子。 注：特殊玻璃仪器都可向玻璃公司定做，可根据提出要求做，可向他们要有关尺寸规格。 2 工艺研究实验室 名 称数 量 台式高速离心机1 低速台式大容量离心机1 超速离心机1 冷冻离心机1 小型粉碎机1 超声波破碎机1 冷冻干燥机1 实验室喷雾干燥机1 微波炉1 冰箱1 分离填料1 高效硅胶预制板2 3 精密分析实验室 名 称数 量 电子分析天平1 紫外分析仪1 标准比色仪1 紫外分光光度计1 食品物性分析仪1 凝胶成像分析仪1 红外水份测定仪1 水份活度仪1
齐晋鞌之战的全文翻译
齐晋鞌之战的全文翻译 1、版本一 晋鞌之战（成公二年） 作者：左丘明 【原文】 楚癸酉，师陈于鞌(1)。邴夏御侯，逢丑父为右②。晋解张御克，郑丘缓为右(3)。侯日：“余姑翦灭此而朝食(4)”。不介马而驰之⑤。克伤于矢，流血及屨２未尽∧?6)，曰：“余病矣(7)！”张侯曰：“自始合(8)，而矢贯余手及肘(9)，余折以御，左轮朱殷(10)，岂敢言病？吾子忍之！”缓曰：“自始合，苟有险，余必下推车，子岂＿识之(11)？然子病矣！”张侯曰：“师之耳目，在吾旗鼓，进退从之。此车一人殿之(12)，可以集事(13)，若之何其以病败君之大事也？擐甲执兵(14)，固即死也(15)；病未及死，吾子勉之(16)！”左并辔(17)，右援枴?鼓(18)。马逸不能止(19)，师从之，师败绩。逐之，三周华不注(20)。 【注释】 ①师：指、晋两国军队。羞：同“鞍”，国地名，在今山东济南西北。②邴(bing)夏：国大夫。侯：顷公。逢丑父：国大夫。右：车右。③解张：晋国大夫，又称张侯。克：即献子，晋国大大，晋军主帅。郑丘缓：晋国大夫，姓郑丘，名缓。(4)姑：暂且。翦灭：消灭。此；指晋军。朝食；吃早饭。⑤不介马：不给马披甲。驰之：驱马追击敌人。(6)未绝鼓音：作战时，主帅亲自掌旗鼓，指挥三军，
所以克受伤后仍然击鼓不停。(7)病：负伤。(8)合：交战。(9)贯：射。穿。肘：胳膊。(10)朱：大红色。殷：深红色。(11)识：知道。 (12) 殿：镇守。(13)集事：成事。(14)擐（huan）：穿上。兵：武器。 (15) 即：就。即死：就死，赴死。(16)勉：努力。(17)并：合在一起。辔（Pei）：马组绳。(18)援：拉过来。枴〉襲）：鼓槌。(19)逸：奔跑，狂奔。(20)周：环绕华不注：山名，在今山东济南东北。 【译文】 六月十七日，国和晋国的军队在鞌摆开了阵势。邴夏为顷公驾车，逢丑父担任车右。晋国解张为卻克驾车，郑丘缓担任车右。顷公说：“我暂且先消灭了这些敌人再吃早饭。”军没有给马披甲就驱车进击晋军。卻克被箭射伤，血流到鞋子上，但他一直没有停止击鼓，并说：“我受伤了！”解张说：“从开始交战，我的手和胳膊就被箭射穿了，我折断了箭，继续驾车，左边的车轮因被血染成了深红色，哪里敢说受了伤？您还是忍住吧？”郑丘缓说：“从开始交战，只要遇到险阻，我一定要下去推车，您哪里知道这些？可是您却受伤了！”解张说：“我们的旗帜和战鼓是军队的耳目，军队进攻和后撤都听从旗鼓指挥。这辆战车只要一个人镇守，就可以成功，怎么能因为负了伤而败坏国君的大事呢？穿上铠甲，拿起武器，本来就是去赴死；受伤不到死的地步，您要奋力而为啊！”解张左手把马绳全部握在一起，右手拿过鼓槌来击鼓。战马狂奔不已，晋军跟著主帅的车前进，军大败，晋军追击军，围著华不注山追了三圈。 2、版本二
齐晋鞌之战原文和译文
创作编号：BG7531400019813488897SX 创作者：别如克* 鞌之战选自《左传》又名《鞍之战》原文：楚癸酉，师陈于鞌(1)。邴夏御侯，逢丑父为右②。晋解张御克，郑丘缓为右(3)。侯日：“余姑翦灭此而朝食(4)”。不介马而驰之⑤。克伤于矢，流血及屦2 未尽∧?6)，曰：“余病矣(7)！”张侯曰：“自始合(8)，而矢贯余手及肘(9)，余折以御，左轮朱殷(10)，岂敢言病？吾子忍之！”缓曰：“自始合，苟有险，余必下推车，子岂＿识之(11)？然子病矣！”张侯曰：“师之耳目，在吾旗鼓，进退从之。此车一人殿之(12)，可以集事(13)，若之何其以病败君之大事也？擐甲执兵(14)，固即死也(15)；病未及死，吾子勉之(16)！”左并辔(17) ，右援拐?鼓(18)。马逸不能止(19)，师从之，师败绩。逐之，三周华不注(20) 韩厥梦子舆谓己曰：“旦辟左右！”故中御而从齐侯。邴夏曰：“射其御者，君子也。”公曰：“谓之君子而射之，非礼也。”射其左，越于车下；射其右，毙于车中。綦毋张丧车，从韩厥，曰：“请寓乘。”从左右，皆肘之，使立于后。韩厥俛，定其右。逢丑父与公易位。将及华泉，骖絓于木而止。丑父寝于轏中，蛇出于其下，以肱击之，伤而匿之，故不能推车而及。韩厥执絷马前，再拜稽首，奉觞加璧以进，曰：“寡君使群臣为鲁、卫请，曰：‘无令舆师陷入君地。’下臣不幸，属当戎行，无所逃隐。且惧奔辟而忝两君，臣辱戎士，敢告不敏，摄官承乏。” 丑父使公下，如华泉取饮。郑周父御佐车，宛茷为右，载齐侯以免。韩厥献丑父，郤献子将戮之。呼曰：“自今无有代其君任患者，有一于此，将为戮乎？”郤子曰：“人不难以死免其君，我戮之不祥。赦之，以劝事君者。”乃免之。译文1：在癸酉这天，双方的军队在鞌这个地方摆开了阵势。齐国一方是邴夏为齐侯赶车，逢丑父当车右。晋军一方是解张为主帅郤克赶车，郑丘缓当车右。齐侯说:“我姑且消灭了这些人再吃早饭。”不给马披甲就冲向了晋军。郤克被箭射伤，血流到了鞋上，但是仍不停止擂鼓继续指挥战斗。他说:“我受重伤了。”解张说:“从一开始接战，一只箭就射穿了我的手和肘，左边的车轮都被我的血染成了黑红色，我哪敢说受伤？您忍着点吧！”郑丘缓说:“从一开始接战，如果遇到道路不平的地方，我必定（冒着生命危险）下去推车，您难道了解这些吗？不过，您真是受重伤了。”daier 解张说:“军队的耳朵和眼睛，都集中在我们的战旗和鼓声，前进后退都要听从它。这辆车上还有一个人镇守住它，战事就可以成功。为什么为了伤痛而败坏国君的大事呢？身披盔甲，手执武器，本来就是去走向死亡，伤痛还没到死的地步，您还是尽力而为吧。”一边说，一边用左手把右手的缰绳攥在一起，用空出的右手抓过郤克手中的鼓棰就擂起鼓来。（由于一手控马，）马飞快奔跑而不能停止，晋军队伍跟着指挥车冲上去，把齐军打得打败。晋军随即追赶齐军，三次围绕着华不注山奔跑。韩厥梦见他去世的父亲对他说:“明天早晨作战时要避开
混凝土实验室建设及仪器的配置方案样本
混凝土实验室建设与仪器的选配 文摘: 本文主要论述了高性能混凝土实验室应具备的检测试验能力, 以及仪器设备配置和实验室环境测量控制等方面的问题。初步探讨了实验室管理的有关问题。 关键词: 混凝土实验室管理高性能混凝土 0 前言 高性能混凝土实验室既是科研机构从事科学研究也是预拌商品混凝土生产企业进行产品研发与生产质量控制的必备基础性条件。高性能混凝土实验室建设与管理的目标是为获得满足科学研究与生产质量控制要求的必要种类的高质量的检测试验结果数据。因此, 对高性能混凝土实验室的建设与管理探讨也应从以上两个方面入手。 0.1关于实验室检测试验数据质量 无论是科学研究还是生产质量控制, 都要求检测试验数据具备以下四种特性或其质量满足以下四个方面的要求: ■真实性; ■准确性; ■代表性; ■完整性。 0.2关于高性能混凝土实验室检测试验能力 高性能混凝土实验室检测试验能力应满足科研机构从事科学研究或预拌商品混凝土生产企业进行产品研发与生产质量控制要求。 1．指导高性能混凝土实验室建设与管理的有关标准 主要管理标准: 检测和校准实验室能力的通用要求 GB/T27025— /ISO/IEC17025:
实验室资质认定评审准则 测量管理体系测量过程和测量设备的要求 GB/T19022- /ISO10012: 用于检测实验室的测量设备的校准间隔确定的指南 D10 主要技术标准: 建筑工程施工质量验收统一标准 GB50300- 混凝土结构施工质量验收规范 GB50204- 混凝土结构设计规范 GB50010- 混凝土结构耐久性设计规范 GB/T50476- 普通混凝土长期性能和耐久性能试验方法标准 GB/T50082- 混凝土耐久性检验评定标准 JGJ/T193- 混凝土结构耐久性评定标准 CECS220: 高性能混凝土应用技术规程 CECS207: 混凝土结构耐久性设计与施工指南 CCES 01- 民用建筑工程室内环境污染控制规范GB50325- 2．高性能混凝土实验室应具备的检测试验能力 高性能混凝土实验室的检测试验能力应满足以下要求: ■有关验收规范标准、设计规范对混凝土及原材料质量进行检测试验的要求; ■混凝土原材料优选与质量控制对检测试验的要求; ■混凝土配合比设计对检测试验的要求; ■混凝土性能检验对检测试验的要求。 因此, 高性能混凝土实验室应具备以下几方面检测试验能力: