轴承退磁处理中的模糊神经控制仿真研究

A Simulation on Fuzzy-Neural Control during the Demagnetization Treatment of Bearing

基于模糊神经控制的方法,选择轴承的残磁大小、检测距离与轴承套圈的直径大小为系统的控制输入变量,应用MATLAB软件建立退磁机交变电流的自适应模糊推理控制器,根据80组实验数据值进行网络训练,自动生成各控制变量的隶属函数分布与模糊控制规则等。通过对退磁机交变电流的模糊神经控制系统仿真表明,基于模糊神经控制的退磁机交变电流控制,能得到很好的退磁机交变电流控制效果,从而实现轴承残磁余量的理想控制,解决了采用经典控制时因很难建立精确数学模型而无法控制的难题,为轴承自动化生产中残磁余量的有效、快速退磁提供了一种新的理论方法。

As the size of bearing residual magnetic detection distance and the diameter of bearing ring to control input variables of the control system,training neural network according to 80 experimental data values in FIS established by ANFIS in MATLAB,generating automatically distribution of every control variables membership functions and fuzzy control rules,etc.The result of simulation on demagnetization alternating current fuzzy-neural control system shows that alternating current control on demagnetization can be well controlled by using fuzzy-neural control method,and it realizes that bearing residual magnetic quality control can get a good control.It solves difficulties of establishing accurate mathematical models by using classical control and provides a new theoretical method for effective and fast control of bearings residual magnetic margin in automated production.