Study On Of Fuzzy Systems And Response Capability Based On The Universal Triple I Method

As a class of approximate reasoning methods dealing with uncertainty, fuzzyreasoning is the critical rationale of the fields such as fuzzy expert systems, fuzzycontrol systems and etc. Study on response functions of fuzzy systems based onfuzzy reasoning is a major research direction of fuzzy control theory. The triple Imethod is a newly-established fuzzy reasoning method put forward by ProfessorWang Guojun in1999. The new method is an improvement of the CRI method.The triple I method and its applications to fuzzy control have witnessedsubstantive progress in theory and practice. But many further studies on them canstill be done.The main research work and results of this paper are as follows:1. Research on the construction and typical characteristics of the universal tripleI method. On the basis of the traditional triple I method, we consider threedifferent fuzzy implication operators is conducted. The triple I method based onthree selected operators is called the universal triple I method for short. Research onthe continuity and approximation of the universal triple I method is also done. Andthe features of error propagation and the influence on the output of fuzzy systemwhen the rules are disturbed are taken into consideration.2. Research on the response function of fuzzy controllers based on the universaltriple I method. We apply the universal triple I method to fuzzy controller withfuzzification, fuzzy reasoning and defuzzification, and output of the control systemcalled the response function is obtained. Based on the universal triple I method,research on response functions of single input&single output and double input&single output under six different situations are conducted. The research result ofthis paper indicates that some fuzzy systems constructed on the universal triple I method possess interpolation and universal approximation, and these system canbe considered to be used in the practical fuzzy control system.3. Research on the influence of change of aggregating rules on fuzzy system’sresponse function. Study on the influences of changing the rule synthesis way―or‖into―and‖on the fuzzy system’s response function. The research result indicatesthat the fuzzy system has step function response and possess universalapproximation. The system can be considered to be used in the practical fuzzycontrol system.4. Simulation of fuzzy controller. Analysis of reasoning rendering of theuniversal triple I method presented in fuzzy controller and approximation of thecontroller are conducted. Finally a self-adaptive fuzzy PID controller isconstructed. Compared with the traditional PID controller, the newly-constructedcontroller has advantages in restraining over-control, the time period of the systemturning into stable state.