The Dynamic Fuzzy Method

With the development of science and society, intelligent control will be demanded widely in the future. Fuzzy control, expert control and neural network control are known as the prospective three-intelligent controls.This paper analyzes the principle characteristic of PID control and the two-dimensional fuzzy control fully, and absorbs the two's merit to propose M-type control. This control is essentially one kind of three-dimensional fuzzy controls. It is three-dimensional fuzzy control with adjustment factors, and better than traditional three-dimensional fuzzy control. It breaks plague of the three-dimensional fuzzy control dimension. For the use of the accumulation, it carries on integral to the deviation, can eliminate static difference of system and increases system's stable state precision. Dynamic system simulation proves superiority of the M-type control.M-type control and the fuzzy control need to establish fuzzy rules, but fuzzy rules'establishment needs site operators' experience and experts'knowledge to be summarized. This is a long-term process. Specially facing a brand-new domain, the experience has not been possible summarized. For this reason, this paper also proposes degenerated M-type control. It does not need to establish fuzzy rules and avoids difficulty of establishment rules. The degenerated M-type controller depends on proportion of the deviation, deviation change, deviation's deviation change to control plant online. The degenerated M-type control is self-adaptive control with identification-free.