论文部分内容阅读
提出了一种基于遗传算法的神经网络自适应控制方法．该方法是针对ＢＰ算法训练神经网络控制系统时收敛速度慢、动态特性不够理想等不足，用改进的遗传算法来优化神经网络辨识器与控制器的参数，以提高控制系统的性能．仿真实验表明该控制器对于非线性、时变、滞后等对象都具有很好的控制精度、鲁棒性和动态特性
A neural network adaptive control method based on genetic algorithm is proposed. This method is aimed at the BP neural network control system training slow convergence, the dynamic characteristics of less than ideal, and so on, with improved genetic algorithm to optimize neural network identifier and controller parameters in order to improve the performance of the control system. Simulation results show that the controller has good control precision, robustness and dynamic characteristics for nonlinear, time-varying, hysteresis and other objects