Fuzzy Logic Controller For Double Inverted Pendulum Using Genetic Algorithms

GA,which is short for Genetic Algorithm,is a global optimization algorithm.GA is applied to mathematical calculations by studying the law of nature,called “the survival of the fittest”.Based on the characteristics of genetic algorithm,such as adaptive,high parallelism and robustness,genetic algorithm is especially welcomed in fuzzy control system to optimize the paramet ers.But the genetic algorithm still has shortcomings,such as premature convergence,poor local search efficiency and so on.Aiming at the shortage of basic genetic algorithm,the author has done the following work: the first one,introducing two indexes who can influence the crossover probability and mutation probability;the second,by analyzeing the given indexes,we will get the corresponding crossover probability and mutation probability increment table,then act on dynamic change of the crossover probability and mutation probability;finally,we will have the adaptive adjustment of crossover and mutation probability.By comparing the result in typical test function with other methods,it is proved that the new algorithm is an effective global search algorithm.The fuzzy controller depend s on people's experience,not upon the specific mathematical model.The fuzzy controller is widely used in the control field with the complex nonlinear system.When the fuzzy control rules are determined,the membership functions of the fuzzy variable play an important role in the control effect of the fuzzy controller.However,the determination of membership function generally requires the method of trial and error,so it usually can not get the ideal control effect.In this paper,the improved genetic algorithm is used to optimize the membership functions of each linguistic variables in the fuzzy controller,and then to control the double inverted pendulum system.The results of Matlab simulation experiment show that the algorithm is effective.