Hybrid Fuzzy PID Control Based On GA

Control system, PID controller algorithm because of its simplicity, robustness, high reliability, simple structure, etc. have been widely used, where PID control key question is PID parameter tuning. However, the traditional PID design methods in the face of increasingly complex nonlinear, time-varying parameters, can’t establish a precise mathematical model and other control objects, it is difficult to obtain good control effect. With the development of the theory of intelligent control, there have been many intelligent algorithms, for example, GA, Artificial Neural Networks, Fuzzy Control, Ant Colony Algorithm, Simulated annealing algorithm, Immune Algorithm. All kinds of intelligent algorithms with conventional PID control methods to organically combine, play to their strengths, but also the intelligent control of the main directions of research.On the determination of the fuzzy rules for fuzzy control bottleneck, the point of this article in the third chapter speaking expert fuzzy rule of thumb is abstracted into a sequence, its precise to draw precise control rule table. Establish a simulated annealing genetic algorithm optimization, drawn optimized control rule, to enable them to meet the integrity, consistency and accuracy of the rule table. Chapter based on satisfaction of fuzzy adaptive PID control method which Satisfactory Optimization system, a very important issue is the establishment of the function of satisfaction, this paper gives an established satisfaction the method of function. Design satisfaction function is then taken as the optimization of the evaluation function, and then using the genetic algorithm to search for a set of PID parameters as the base value, change the value of conventional trial and error, or the group taken in the light of experience, it will be more close to the optimal value, thus shortening the time control, and then adjust the PID parameters online, overall satisfaction function evaluation after each adjustment to get a good control effect by Matlab simulation results proved the feasibility of the designed controller.