The Study Of Low Speed Wire Walking Constant Tension System Based On Fuzzy Neural Network

This dissertation is mainly about the research on the development about low speed wire walking constant tendion system. The ability to control the constant tension is one of the important targets to evaluate the performance of low speed wire walking machine tool, which will directly affect the machining precise and machined surface quality. With the developing needs on micro and high precise manufacture these years, more requests on the stability of wire-feeding system are called on. Therefore, it has profound meanings to research on the wire-feeding system.The new developments of low speed wire walking machine tool and the wire-feeding system for low speed wire walking machine tool are stated first in this paper on basis of lots of domestic and overseas technique literature on low speed wire walking machine tool. After analyzing and researching on the working theory of machine tool's tension system, a method to control close loop tension using DC torque motor control is presented. According to the qualification of the low speed wire walking machine tool's tension control, this paper chooses the model of the wire system components . And it established an accurate mathematical model of the wire system and studied its dynamic characteristics.This dissertation uses the inference Takagi-Sugeno model construct a fuzzy neural network controller. On the basis of constant tension system performance demands, it selects the various parameters of the fuzzy neural network controller and applied fuzzy neural network controller into constant tension control system. At last, it simulates the constant tension system in different error inputs.