Double-Weighted Neural Network Theory And Its Application In Control Field

Double-Weighted neural network is a new type of neural network that put forwarded by WangShouJue in recent years, who is an Academician of Institute of Semiconductors of Chinese Science Academy.The neural network is very comprehensive. BP networks and RBF neural network can be its special case by set up the parameters. Many documentations was advanced on function approach, system identification, mode recognition and so on since Academician Wang put forwarded it. All of the study achieved good results and have demonstrated the high value of the Double-Weighted neural network.This paper will introduc some theory about Double-Weighted neural network-the approach ability of the Double-Weighted tensor product neural network, and study the control capability of the Double-Weighted neural network in control field. The purpose of the paper is to make the application of Double-Weighted neural network further promoted. We will make two experiments with Double-Weighted neural network and BP neural network respectively to identfy the system and control its output to trace the anticipant output . Then, we will compare control performance of the two neural network controllers.The results showed that control performance of Double-Weighted neural network controller is well than BP neural network controller. The main content of the paper is arranged as follow:...