In this paper, using the advantages of single variable neural network and PID controller, the weight values are initialized by the combination of the PID parameters tuning and the existing knowledge of the controlled object. On this basis, the improved single variable neuron closed-loop controller is simple and fast because of the simplified control algorithm. And by the simulation comparison, the proposed method of this article is much better than the method of the conventional single neuron controller.
Huajun Zhang;庆学 黄;Long Quan
Journal of Information and Computational Science
2012-10