The Study Of Fault Diagnosis Algorithm Based On FTA

As more and more large-scale industrial systems, the algorithm about fault diagnosis and analysis of the system has become increasingly complex. Fault Tree Analysis, FTA for short, is gradually being accepted by people as a fault diagnosis algorithm which is simple, effective and promising.In this paper, we build a relatively complete algorithm of fault diagnosis based on FTA, which include two parts. In the achievements algorithm, we have improved the original algorithm which is based on Decision-Table. We make the classification to the reason and make the appropriate rules; thereby we can reduce the query to determine the number of traversing the table. So we can improve the efficiency of the fault tree generation. In data analysis, we use the early non-intersection method to get the fault tree structure function, and make the fault tree of system combined with reliability parameters to get the function about the fault of system and time. At the same time, we offer the degree of component importance to determine priorities for system, and provide a more complete diagnostic program for the system fault.Finally, we demonstrate our conclusion with a continuous stirred tank reactor. In the conclusion, the fault tree is built, while the probability density function about top event has been given. We show the priority of the components that the system run under 1000 hours and compare with the traditional method. It proves that the priority under the components has higher efficiency than traditional programs.