New Approach of Analog Circuit Fault Knowledge Acquisition Base on Simulation

摘要

The difficulties of acquiring fault knowledge severely handicap the development of intelligent diagnosis system (IDS) of electric equipment, especially in analog field. In this paper, the general procedure of knowledge acquisition (KA) by fault simulation is analyzed at first. And then, a new method is presented, which auto-acquires fault knowledge by simulating all possible faults of analog circuit. Here four key strategies are necessary to reduce simulation complexity to a tolerable level and to establish the fault set automatically. The scheme of circuit fault knowledge acquisition flat is proposed to implement this method. Automation of KA, to a certain extent, solves the basic problem of hard task of gathering the fault original data, and makes the knowledge acquired more sufficient. Finally, results of experiment prove its feasibility.