一种基于同伦BP网络记忆模糊规则的参数自调整ANN-PI控制及应用

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,本文设计了一种参数自调整ANN一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.

The Control and Application of a Parameter Self-Adjusting ANN-PI System Based on the Fuzzy Rules Using Homotopy BP Algorithm

Back-Propagation neural network can record the fuzzy control rules efficienity and utilize these experiences according to associative memory. All now unavailahlly feedforward-net learning algorithms have a locl minimum problem. The homolopy continuation BP algorithm pro-vidcs an effective method for global convergence of BP network, and is of very fast convergent speed. To enhance control system's accuracy and disturbance-resisting ability, an ANN-PI contrel ler with the parameters self-adjusting has been designed in the paper. The results of experiment show that this controller has a faster dynamic response, higher control accu disturbance-resisting ability, less sensitive to parameter changes, and robustness.