Building Materials Quality Inspection System Based On IC Card And Barcode And Analysis Of The Experimental Curve

Construction industry is the pillar industry of the country's economy, it is capital and technology-intensive industry, due to China's basic national conditions, the construction industry has monopoly, uniqueness, with nature of permanence, huge scale and so on. The quality of construction is closely related to the used building materials, in order to protect the safety of the person and the property, the quality of the building materials needs strictly checked, for which the state sets quality standards of building materials in details, different provinces and cities have also developed specific quality standards.At present. national and local government is vigorously promoting the use of computer models instead of the traditional manual test to make the quality inspection report scientific, fair and accurate. The original manual management, for example, artificial dial reading and manual filling of the testing data, etc, has been far behind with the rapid development of China's economy, in this context, "the quality of construction materials inspection system" came into being. The first phase of the system's design and implementation has been successfully applied in Qiqihar City Construction Quality Supervision and Inspection Center, the secondary development of the system will embed the original IC card and barcode into the software.This paper analyzes the processes of the quality of the building materials, the existing modules and the data partitions in "the quality of construction materials inspection system" then make a second development, implants the IC card and barcode into it, which not only improve the quality of the original automated inspection system, but also further improve the quality of working safety, effectiveness and convenience. The new IC card module and the new barcode module have complete functions and work independently, the interface of interaction with other modules is simple and effective, and it has strong robustness and encapsulation. IC card module firstly package to get a common user interface, it reads and writes units of 4 bytes, which makes it has a larger amount of information, this kind of usage in building materials quality inspection system is more rational and efficient; Then, when it comes to the specific type of IC SLE4442 card, we use JNI technology to construct a C stub, which establishes one to one bridge between the C and Java operating functions and achieves the mutual communication between the local interface and Java Virtual Machine platform; Finally, we use IC Card embedded function to install the timer to the user interface that needs monitoring. Barcode module simultaneously provides three functions:barcode generating function, after the commission received samples, it will pop-up dialog box of commissioned number, at the same time, call the barcode generating function, you can specify the file path and it will produce barcode file under the path; Report barcode transfer function, it realize the mechanism of report barcode transmission, when add data to the. jasper report files, we can map the commissioned number to barcode, when display or print the report it will generate barcode on the upper right corner; Barcode response function, achieve automatic open of samples' experiment, and automatic storage of reports'issue, the barcode response function module can be embedded in the experiment module and issue module, the method is first add to user interface a text control which can save barcode value, and then call the barcode response function in monitor function.Tensile test is the most important one to determine the mechanical response, any building materials in the tensile test can get a Force-Elongation curve, which is extremely important graphic that describes the mechanical property of the materials, from the Force-Elongation curve, you can get the elastic modulus and the variation of the materials, plastic changes in the materials characteristics and materials strength, ultimate deformation capacity. There is a very special class of curve, which is controlled manually by skilled craftsman, it is very similar with real curve, it is difficult to distinguish using manual methods. In order to differentiate the real from the fake and to guarantee the authenticity of the detection, this paper forwards the analysis of Force-Elongation curve.Time series is a group of data sequences which ordered and arranged according to date. In reality many sequences can be extended to time series, for example, we treat the student number as the time, student marks sorted by the student number is time series. Force-Elongation curve is the generalized time series, we treat the axis of elongation as the axis of time, the force is time-varying data, the analysis of Force-Elongation curve, can be abstracted into the problem of time series classification.There are many real life time series, the value of the adjacent sequence has close trend, it can be approximated to that certain segments have the same trend, and this paper abstracted it as MSL (Multi-Segment Linear) features. This paper proposes a novel time series classification algorithm FEC (Force-Elongation Classification), which consists of three modules:Derivative valuation function, linear segmentation method. DDHMM model (based on the HMM). Derivative estimation function uses the basic principle of replacing the derivative with difference to get the average value of the slope of the time series, the input is time series. the output is the derivative estimation of the sequence value. Linear segmentation method, by calling the derivative valuation function, segment the time series by the trend of the similarity, the output is series after segmentation, the output can be used to determine whether the time series meets the MSL features, if met, time series will be converted into observed sequence with special structure, each of observations consists of the segment-label (it can indicate the basic shape of the segment) and all sequence values of the segment, due to linear segmentation method is affected by the error s, the divided segment is quasilinear. while all sequence values need clustering method to be discrete. in which there will be errors, if we do not treat the value as a unity,the error will be ignored, and the classification results will be affected. DDHMM (Derivative Duration Hidden Markov Model) model analyze the structure of observation, adjust the HMM model parameters, give all new parameters, and the revaluation formula of all parameters. DDHMM can be used to indicate observed sequence with special structure.FEC algorithm basic steps:use derivative estimation function and linear segmentation method to segment the time series, the output is the sequence after the segmentation, the output can be used to detect MSL feature, if met, it will be converted to observed sequence with special structure, then we use the training observed sequence in the training of DDHMM model, after the training is completed, the maximum likelihood decision rule for classification will be used, that is calculating the probability of the testing sequence of observations generated by the model, the model to which the label of testing observed sequence according will output the maximum value of the probability.Result of FEC algorithm experiment shows that:The function of derivative estimation and the method of linear segmentation can be used together to detect MSL feature, and offer the sequence that is segmented; For time series which met MSL features, FEC algorithm will convert it into observed sequence with special structure, the classification through training DDHMM has a high accuracy, and FEC algorithm has a good performance in the classification of materials Force-Elongation curves and the recognition of UCI pen tip trajectories records.