Graphical representation technology for knowledge acquisition of fault diagnosis system

Fault diagnosis expert system (FDES) which can realize high efficient vehicle launch test is an important part of launch protection system in the future space vehicle. The existing way of knowledge acquisition is the artificial extraction rules by knowledge engineers according to the testing process provided by technical personnel. However, this way usually has some problems such as the complexity of the heavy workload, long working cycle, error-prone and does not readily accessible for inspection etc. To solve these problems, this paper presents a method of knowledge acquisition based on graphical representation. Firstly, analysis the testing process of the launch vehicle deeply and establish the overall framework which contains all the preparatory stage subsystem. Then, draw out of the launch vehicle testing process by graphical method through setting reasonable graphical components. This test flow chart is convenient and intuitive to understand, and easier to maintain. Finally, through the design of rule extraction algorithm, the test flow chart is automatically converted to the knowledge which is required by the knowledge base of FDES. This knowledge acquisition method based on graphical techniques can effectively simplify the difficulty, easy to maintain and understand. After testing, this method works well application effect.