基于改进遗传算法的单神经元自适应PID控制器

Adaptive PID Controller of A Neural Network Based on a New Genetic Algorithm

针对现有单个神经元的自适应PID控制器和遗传算法存在的不足 ,提出一种改进的遗传算法 ,并将其应用于单神经元自适应PID控制器中。该方法采用实数编码 ,模糊大变异操作 ,是为了提高收敛速度。仿真结果表明 ,基于改进遗传算法的单神经元自适应PID控制器不仅具有良好的控制性能 。

Aiming at the deficiency and limitation of present adaptive PID controllers of a neural network and genetic algorithm (GA), a new GA is proposed and is applied to the adaptive PID controller of a neural network The coding method of real number and fuzzy big mutation operating can enhance convergency speed The simulation results show that the adaptive PID controller based on the new GA not only has a satisfying control performance but is superior to the conventional adaptive controller in algorithm convergency speed