论文部分内容阅读
文中在分析RBF神经网络整定PID算法优缺点的基础上,给出了一种采用遗传模拟退火算法来优化网络结构和权值参数的RBF神经网络,将改进的RBF神经网络用于整定PID控制,并给出了相应的仿真测试例子。仿真实验结果表明,与采用梯度法优化网络权值等参数的RBF神经网络相比,给出的优化算法能更好地辨识控制系统,具有通用性好、调节精度高、在抑制超调量能力强等优点。
Based on the analysis of the advantages and disadvantages of RBF neural network tuning PID algorithm, a RBF neural network using genetic simulated annealing algorithm to optimize the network structure and weight parameters is presented. The improved RBF neural network is used to tune PID control, And gives the corresponding simulation test example. The simulation results show that compared with the RBF neural network which uses the gradient method to optimize the network weights and other parameters, the proposed optimization algorithm can better identify the control system, has good versatility, high regulation accuracy, Strong and so on.