切削过程的位置误差补偿控制研究

A RESEARCH ON THE POSITION ERROR-COMPENSATING CONTROL OF CUTTING PROCESS

本文作者运用经典控制论对切削过程的位置误差补偿控制进行了研究。接着,作者对神经网络理论中的标准反向传播算法BP作了改进并提出了IBP算法。基于IBP法,作者构造了一个用于切削过程位置误差补偿控制的多层前馈神经网络控制器,并作了仿真控制试验。研究结果表明:IBP法较标准BP法具有更快的迭代速度和收敛性质;所建的智能控制器虽然结构简单,但在误差补偿控制方面跟经典控制法相似具有一定的控制效果。

In this paper, author studies the position error-compensating control problem of cutting process by using the classical control theory. Next, an Improved Back Propagation(IBP)algorithm is proposed, considering the drawbacks of the standard Back Propagation (BP) in the neural network theory. Based upon the IBP, a multilayered feed-forward neural network controller is constructed for the position error-compensating control of cutting process, and a simulation control test is done. The result of research shows: compared with the BP, IBP method has faster iteration speed and better convergence property; although the built intelligent controller has simpler constructure, the control effect is better, so is the classical control in error-compensating control.