Reliability Analysis Of Diesel Engine Based On Fuzzy Fault Tree And Event Tree

With the development of technology and social progress, mechanical system has become more and more complicated, the performance parameter become more and more, working environmental become more and more serious. Under the circumstances, risk analysis is essential for the mechanical system. Many techniques are used in risk analysis in recent years such as Risk matrix, decision tress, Monte Carlo and sensitivity analysis. However some of these techniques are limited because they either do not support quantitative risk analysis, or too difficult to be utilized due to required amount of data. To address such limitation, Engine is chosen as research object, and the fuzzy logic, failure mode and effect analysis, fault trees and event trees are combined. The three methed are integated in the risk analysis instead of one, all this make risk analysis more accurate. The main work is as follows:(1) FMEA and fuzzy logic are combined to assess risk. In the fuzzy FMEA, linguistic terms are used to assess the Occurrence, Severity, Detection of FMEA, meanwhile proper membership function and fuzzy logic rule are built. Fuzzy expert system is build using MATLAB fuzzy box. This entire make it available when there is not enough date. The build fuzzy expert system is used to assessment the five risks of engine, and risk prioritys are obtained. The contrast of fuzzy FMEA and traditional FMEA are analyzed.(2)According to the results obtunded from fuzzy FMEA, more analysis is done to find the cause of failure using fuzzy fault tree analysis. In the fuzzy fault tree analysis, the linguistic terms are used to assess the probability of occurrence. The quantitative and qualitative analyses are done in the fuzzy FTA, and the fuzzy probability of top event and fuzzy importances of basic event are obtunded. The basic event importance ranks are divided to distinguish the influence of basic event.(3) On the basis of the above analysis, the fuzzy logic is combined with event tree analysis, and the fuzzy event tree analysis is constructed. Linguistic terms are used to evaluate the probability of occurrence and impact. The build fuzzy event tree analysis can make it available when there is not enough data. Mitigation strategies are put forward to the risk event. The expected risk magnitudes are obtened under the mitigation strategies. This can make it intuitive and accurate in effect judgment of mitigation. The fuzzy FTA is used in the fuzzy ETA, and the results of the two methed are support each other.