基于改进BP神经网络的智能控制方法研究

Investigation of method for intelligent control based on neural network

针对典型 ＢＰ神经网络存在的缺陷而提出了一些有效的改进措施。通过采用改进的 ＢＰ神经网络来对控制规则样本采样的学习和训练，使网络记忆控制规则，以达到智能控制的目的。仿真和实验结果证明该方法具有优良的控制特性，满足伺服电机控制的需要。

According to the defect of the classical BP neural network, some effectively improved method are presented. Then through the improved BP network to study and practise in the control regulation of sample, the control regulation are remembered by the network to realize the intelligent control. The simulation and test results show the method has excellent control characteristics and strong robustness, which meet servo motor control needs.