铝包钢丝中频加热的模糊神经网络控制技术

在铝包钢丝生产过程中,采用在线中频感应加热方式。基于中频感应加热条件,针对钢丝直径、钢丝运行速度、加热温度、中频频率、中频电压、中频电流、中频功率、加热炉长度之间的函数关系,建立了连续包覆过程的多变量神经网络控制方法。经过多次实验和筛选,选出有效数据,多个模式反复学习,直至网络全局误差函数E小于预先设定的一个极小值。通过网络训练和仿真分析得到较佳的解决方法,该方法应用在铝包钢丝连续包覆生产线上,较好地解决了钢丝加热控制难的问题。

Fuzzy Neural Network Control Technology of Aluminum-Clad Steel Wire IF Induction Heating

In the productive process of the aluminum-clad steel wire, the online IF induction heating method was applied. Based on the IF induction heating condition and aiming at the wire diameter, the wire speed, the heating temperature, the IF frequency, the IF voltage, the IF current, the IF power, the functional relationship between the length of heating furnace, the multi-variable fuzzy neural network control method of a continuous coating process was set up. After doing several experiments, screening to select the valid data, and studying repeatedly in a number of patterns, finally the network universal error function E was less than a pre-set extremely lows. Through the network training and simulation analysis, the preferable method was figured out. The method was used in the production line of a continuous coating of aluminumclad, and the problem of steel wire heating control was preferably solved.