一种基于遗传算法的单神经元PID控制器参数优化

Parameter Optimization of Single-neuron PID Controller Based on Genetic Algorithm

文章分析了神经元PID控制器存在的问题,提出了一种用遗传算法对单神经元控制器进行参数寻优,并利用种群数N和交换概率Pc逐次递减的方法来提高计算效率和收敛速度,在搜索空间内获得全局最优点的方法。试验仿真结果表明,基于遗传算法的单神经元PID控制器参数优化方法能获得很好的控制效果。

This paper analyzes the existing problems of single-neuron PID controller, and puts forward a new type of genetic algorithm to optimize parameters of single-neuron PID controller in the completely search space, and uses the method of gradually reducing the population number N and crossover probability Pc to improve the computational efficiency and convergent rate. It is showed that optimization result of the parameter of single-neuron PID controller based on genetic algorithm can get a good result by way of simulation.