RBF神经网络PID控制在分切机中的应用

PID Control Based on RBF Neural Network and Its Application in Slitter

BOPP生产线中的分切机是一个复杂的机电控制系统,分切机中的张力控制是影响生产质量的主要因素。分析了BOPP生产线中的分切机张力产生原因和张力控制的数学模型,针对分切机的恒张力控制,提出了一种基于RBF神经网络的改进PID控制方法。该方法通过RBF神经网络的Jacobian信息辨识,结合增量式PID算法对张力实现控制参数的自整定,并运用Matlab软件编程实现整个模型的控制仿真。仿真结果表明,改进后的控制算法比一般的PID控制具有更好的控制效果。

The slitter in biaxially oriented polypropylene （BOPP ） production line is a complex mechanical and electric control system, and the tension control of slitter is the main factor to impact the quality of production. The reason leading to tension and the mathematical model of tension control are analyzed. Aiming at the constant tension control for slitter, an improved PID control method based on RBF neural network is proposed. Through Jabcobian information identification of RBF neural network combining with incremental PID algorithm, the self-tuning of parameters for tension control is implemented. The control simulation of entire model is realized by adopting Matlab software programming. The result of simulation indicates that the improved control algorithm offers better control effects than conventional PID control.