摘要
文章提出了一种新颖的 PI-神经网络混合速度控制器的方案。在该方案中 PI参数是不变的 ,控制器的学习由隐含层和输出层之间的权值自调整来实现 ,而且 PI参数可在一个较大的范围内取值。
A novel scheme of speed controller based on hybrid PI controller and neural networks is presented. In this scheme, the parameters of PI controller are constant, and learning is realized by auto-tuning of connecti v e weights between the hidden and the output neurons. The parameters of PI contro ller are selected in a larger range. The simulation results verify that the sche me is feasible and valid.
出处
《电气传动》
北大核心
2004年第1期23-25,共3页
Electric Drive
基金
武汉青年科技晨光计划资助2 0 0 0 50 0 40 2 7
