Optimal Research Of Fuzzy Control Rules

As a kind of intelligentize control, fuzzy control has been used to solve many real control problems comprehensively in recent years. Especially it is often used to control some complex course of industries and some systems which are non-linear or uncertain or even can't be made to build exact mathematic models, and the results are very precise. Fuzzy control is based on the experience and information of experts, so the only things that need to do is to grasp the experience from operators and experts. It is very different from methods of traditional controller which are based on the mathematical model of the plant. In fuzzy control systems, if fuzzy means and fuzzy decision means have been chosen, the core question of control lies in fuzzy control rules.In fact, the fuzzy control rules are based on the experience and information of operators and experts. However, it is difficulty to obtain perfect fuzzy control rules in some complicated industries process, so that fuzzy control rules are absence or roughness, and the control effects will be infected, if the parameters of control objects are changed. From some means, fuzzy control rules are good or not will directly affect capability and character of the control systems. So to optimize the fuzzy rules of control systems is very important for improving the system capability and character The leading content at this paper are outlined as follows: 1. To introduce the background, meaning and the development of fuzzy control, the significance of optimizing fuzzy control rules in fuzzy control system. 2. To analyze the effect to fuzzy control systems performance which is stemed from the quantification factor, to optimize the fuzzy control rules by adjusting the quantification factor, so performance of the control systems will be improved through adjusting the quantification factor .The software MATLAB is used to simulate one order system, two orders system and three orders system, from the results, a important conclusion can be deserved : the fuzzy control rules can be optimized by changing quantification factor. 3. Combining the fuzzy control and the neural, the method of clustering which is used commonly is analyzed, and its' shortage is found out. In terms of the finding, some solutions are put forward. Comparing the former method with the method that is changed, by the simulation software MATLAB, the conclusion is that the later is more precise. 4. The relation between approach precision, number of control rules and the degree of complexion of the control system are analyzed: the approach precision rises as increasing the number of fuzzy control rules, but at the same time the control system's complexion degree will be enhanced. They are conflict, so this paper brings forward to optimize the control rules by amending the border of approach precision, at the same time the complexion degree of control system is simplified. The paper gives a few of examples to illustrate that the approach precision that is needed can be achieved only through using lesser control rules and the system is more simple.