The Improvement Of Fuzzy Controller Based On Rules

The paper discussed one of international and domestic academic hot spots - the fuzzy control of intelligence control field, and proposed a kind of new fuzzy controller– varying the universe and switch rules fuzzy control. Having carried on the beneficial exploration in the speeding up of response and the realization without the over adjusting, and having established a new fuzzy controller based on it.This article designs the content as followed:1. In the usual fuzzy control system, considering the fuzzy controller simplicity, all is taking the system error and the erroneous change as the input, therefore it is similar to the conventional controller of PD. This kind of fuzzy controller has the over adjusting.In order to improve the fuzzy controller's characteristic, we had analyzed the influence of controller variable universe to the fuzzy control's response rapidity and the over adjusting.2. Looking from the interpolation theory, the fuzzy control may be described in the interpolating function F(X). It is equivalent with its ideal response function summarizing the control rules with the interpolating function f(x).Responding function is full approaching the real response function or not need the peak between two distance smaller. This means the control rules must be quite many, and this is very difficult because of the fuzzy controller depending expert knowledge..Based on interpolation viewpoint, this article added the control thought changing universe in fuzzy controller. It related quantification factor error with the output, and also realized fuzzy control system's universe compressing (the peak changing to smaller) or expanding (the peak distance changing in a bigger way). Correspondingly, the fuzzy control precision and the reaction rate changed along with the output changing.Thus we enabled the fuzzy controller's response rapidity speeding up and have realized the non- over adjusting response.3. Analyzing from changing rules.The former fuzzy controller is that once we have defined the rules, which only is carried on the adjusting in the single rule. This article is that reacting is changing according to the output changing, then rules also corresponding changing.Thus we can make the system responding time speeding up and also realize the system faster to achieve balance.