Research on Optimum Model for the Best Link of Expert System

According to the observation of known facts, researcher finds all performances which are conform to the observation of cognitive to associate with form knowledge base, and then build an initial network model. Based on this model, we complete the mutual uncertainty constraint by using the new achieve probability and effect probability between evidences and adjacent nodes. And on this basis, using the drive method which satisfied competition threshold, keeping reasoning to the high level, until find an object conclusion meets the expectation. The confirming result makes the best assuming evidence become the centre of focus from competition. It decides the orientation of evidence which is priority to be chosen, and give a high efficiency solution for choosing deterministic target by rapid positioning and accurate selection.