采用GA提高模糊控制器适应性的方法及应用研究

对于工业控制中广泛存在的多容量、大惯性、大延迟的被控对象,模糊控制是一种较有效的方法,模糊控制中隶属函数的正确选择是模糊控制器设计的关键,针对传统的获取隶属度函数方法的不足,本文设计了一种基于遗传算法的模糊控制器,系统采用遗传算法优化模糊控制器的隶属函数及其量化因子和比例因子的初值.仿真结果表明,采用该算法设计的模糊控制器较传统的PID控制器与模糊控制器具有鲁棒性强、超调量小的特点.这种改进型的模糊控制器具有良好的控制性能,体现了遗传算法在参数寻优方面的优越性.

Improving adaptability of fuzzy logic controller using genetic algorithms and its application

Based on the characteristics of the multi-capacity,large-inertia and delay in the industrial control,fuzzy control has been proven to be effective.In fact,the correct choice of membership functions in fuzzy control is the key to fuzzy controller's design.To improve the limitation of the traditional method of getting membership functions,a fuzzy logic controller(FLC)based on genetic algorithm is proposed.GA is employed to optimize the membership functions,and the initial value of the quantifying factors and scaling factors.The simulation result shows that the new controller represents strong robustness and small overshoot compared with fuzzy FLC and traditional PID controller.The control performance of this fuzzy controller is favorable.The advantage of the genetic algorithms is indicated especially in the parameter optimization.