Surface To Air Missile Intelligent Fault Diagnosis Expert System For Key Technologies

This paper studied on several key problems of Fault Detection Expert System on ground-to-air missiles.The fly test of ground-to-air missiles is a course, including first guide period, middle guide period and final guide period, which is a distinct feature that we applied in the building of our Fault Detection Expert System. In the main structure we adopted multi computers communication frame and connected expert models based on rules, neural net and faults emulation to a integrated system .we expressed our domain knowledge or experts knowledge in the forms of rules, coefficients of weights in neural net and faults emulation models. Based on the character of course we built our reasoning and control strategy. On the trunk line of reasoning we reasoned based on rules, then on the subtle line of reasoning we reasoned with one of the three inferential models: models based on rules, neural net and faults emulation. In this paper many detail inferential flow charts were supplied. Considering of the system抯 character, we designed the explanation program by presetting commendatory and inferential flow trace.We built a simple demo Fault Detection Expert System on ground-to-air missiles. This system has simple man-machine conversation, explanation and fault detection. We also built a knowledge base management program. A table of fault detection results was given and one detailed example ware demonstrated in this paper. The result of this system shows that the main design of the Fault Detection Expert System on ground-to-air missiles is viable and effective.