Reliability Analysis Of Automobile Detection System Based On FTA

At the present, with the development of the automotive industry and the number of vehicles in total increasing, there are more and more vehicles waiting for detecting, in order to increase the speed of inspection and testing of vehicles, it is requested that reliability of detection lines must be higher. In addition, following the increase in vehicles on the road, traffic safety is increasingly serious, and environmental issues have also become the focus of attention. Therefore automotive safety testing, exhaust and noise testing has been more concerned in the transportation sector, and the reliability of automobile detection line is appealed more closely, so that the study of the automobile detection reliability has not been delayed. Now, the study of the automobile detection reliability at home and abroad is scarce, while the study of reliability in other fields has been widely used, so there is great practical significance doing the research on the Automobile Detection System.The reliability analysis is as seven Automobile Detection lines of northeast as the research object in this article, launching on the Automobile Detection system reliability analysis. The main purpose is that making statistics of the common fault from the Automobile Detection System, based on which analysis of the fault cause is done, and setting up system fault tree, and then do some qualitative and quantitative analysis so as to find out the reasons for the failure in combination, and calculate probability and the fuzzy probability importance degree of system modules, accordingly to ascertain weak parts of the Automobile Detection System. By making management, maintenance, optimization or improving for the weak parts of the automobile detection system, the occurrence of faults hidden dangers is reduced. Finally, according to the characteristics of fault diagnosis, the article establishes the fault diagnosis expert system based on production rules. Once the detection system did not work regularly, staff could find the cause of the malfunction and remove methods timely through the fault diagnosis expert system. Not only improving the efficiency of the detection system, but also alleviating the inconvenience of the expert or maintenance staff repairing personally to the defection line.The Automobile Detection System is integrated by machinery, electronics, automatic control and network, and it is a complex mechanical-electrical integration system. Failure modes are divery, cause of the malfunction is much, so it is difficult to determine failure probability accurately. In view of the above problem, this article adopts to FTA based on fuzzy logic. In this article, the core content is as followings:1. The fault data collection and the establishment of reliability database Before the fault analysis,for the need of reliability study and fault tree analysis ,fault dates should be investigated by going deep into automobile detection line,adopting to actual fault collection methods. In the fault dates collection foundation,combining the macroscopic constitution of automobile detection system(The automobile check-out facility system, the automobile examination observation system and the network communication and the information management system),faults are recognized and anal sized. In this article, select 18 regular failure sites of the three subsystems in a total,collect 72 common failure models,and analyze 180 fault cause corresponding to the 72 failure models. In view of the above findings, the establishment of the automobile detection system troubleshooting database.2. Fuzzy mathematics theory of fault tree analysis The FTA of the Automobile Detection System will use fuzzy mathematics theory. The article needs to identify the fuzzy probability of the basic events, the fuzzy probability and fuzzy important degree of the failure mode (sub event) in the FTA of the Automobile Detection System. The occurrence probability of the basic events gained by various ways has many forms, including accurate occurrence probability value, language value and various fuzzy numbers. Therefore, for the sake of fault analysis, they should be unified. Because trapezoidal fuzzy numbers are membership function of linear distribution, algebra operations between them are simple, and translating other forms of fuzzy values into trapezoidal fuzzy numbers is easy and intuitionist, so various fuzzy values are translated into trapezoidal fuzzy numbers. In the course of quantitative analysis of the FTA, fuzzy operator is used for fuzzy operators and relationship operations of the fuzzy numbers.3. Fault tree analysis based on fuzzy logicAccording to the fault collection, analyses logic relationship of the fault events, and determine the top events and events at all levels of the fault tree, in order to express the logic relationship of the fault events at all levels, they are linked by the specialized symbols, that is establishment of fault tree of the Automobile Detection System. On the basis of fault tree, minimal cut sets which affect reliability of the automobile detection system are calculated.Making quantitative analysis to fault tree, this article applied various fuzzy numbers, language value and precise probability value to describe occurrence probability of fault events. Overcome difficulties conventional fault tree met with while analyzing complex system, at the same time did not exclude conventional FTA method, regarded it as special case of fuzzy fault analysis, therefore made fuzzy fault tree analysis have wider applying range, showed its'big agility and applicability. Various fuzzy numbers were used to evaluate basic events, and fuzzy set theory was applied to analyze fuzzy fault probability of every sub events, and then made fuzzy operations with fuzzy fault probability of sub events, so that got the fuzzy fault probability of top event. Meanwhile centroid of area method was used to calculate fuzzy importance degree of sub events, accordingly to ascertain key parts and weakness parts. The fuzzy fault tree analysis method that put forward in this paper has big practicability and feasibility, and it is a good way for complex system reliability analysis.4. Improvement of the Automobile Detection System reliability.Based on conclusions of the Automobile Detection System fault tree analysis, making some reliability improvements on certain aspects, such as equipment procurement and transport, plant design and construction, installation of equipment and using of system, maintenance and software of the Automobile Detection System.5. Research and Development of the Fault Diagnosis Expert SystemIn view of the issue that structure of the Automobile Detection is complex and maintenance is very difficult,the article has set up an Fault Diagnosis Expert System of the Automobile Detection System. According to the characteristics of fault diagnosis, the article establishes the fault diagnosis expert system based on production rules,and in use of forward chaining inference method,so as to achieve fault diagnosis and troubleshooting. In addition, in order to facilitate user, friendly and intuitive man-machine interface is designed, and the article demonstrates the system's operational procedures using examples.