论文部分内容阅读
BP神经网络可有效地记忆模糊控制规则,并以“联想记忆”方式使用这些经验。然而,现有前向网络学习算法不可避免地存在局部极小问题,同伦连续BP算法可有效地解决BP网络的全局收敛性问题,同时使网络具有很快的收敛速度。为了进一步提高控制系统的精度和抗干扰能力,设计了一种参数自调整ANN-PI控制器。实验结果表明,这种控制器动态响应快,控制精度高,抗干扰能力强,对参数变化不敏感,具有一定的鲁棒性。
BP neural network can effectively remember fuzzy control rules and use these experiences in “associative memory” mode. However, the existing forward learning algorithm inevitably has local minima. The homotopy continuous BP algorithm can effectively solve the global convergence problem of BP network and make the network converge quickly. In order to further improve the accuracy and anti-interference ability of the control system, a parameter self-adjusting ANN-PI controller is designed. Experimental results show that this controller has some advantages such as fast dynamic response, high control accuracy, strong anti-interference ability, insensitivity to parameter changes and certain robustness.