SELF-LEARNING ALGORITHM OF FUZZY SEMANTIC INFERENCE
Q Uang, Hang Zhao, Z. Shen · Acta Automatica Sinica · 1984
For the control of complicated nonlinear systems, fuzzy semantic inference has its great superiority. However, since the objective processes are very complicated,it is difficult to exactly provide the correspoding relationship between the conditional semanteme and conclusive semanteme in advance. For practical application, it is desirable for this relationship to be obtained by self-adjusting or even self-generating in the control system. Furthermore, because the weight inference exerted by the inference consequence of multi-factor input on the final consequence varies at random, the regularity of the weight variance can hardly be observed in a large and complicated system. Therefore, this paper presents an algorithm solving the self-generating of this regularity in a control system with complicated processes.