基于Adaline的数控系统位置控制器研究

Study on feed position control of CNCS based upon Adaline

针对机床数控系统进给位置控制的性能要求,必须根据其工作方式和误差情况实行变增益控制策略,才能较好地兼顾动态和静态指标,为此,选用自适应线性神经元(Adaline)代替原有的比例环节,可以获得相应的控制性能。经数字仿真和实际使用表明,在数控系统中引入人工神经网络控制思想是合理可行和有效的,并且还可以进一步通过硬化实现,达到实时在线控制的目的。

According to the performance demands of the feed position control in the Computer Numerical Control System（CNCS）, the adaptive gain regulation control strategy must be accomplished by uso of the position error, so that the satisfactory balance of the dynamics and static performance is achieved. Adopts the Adaptive Linear Element（Adaline）of Artificial Neural Network（ANN）, instead of traditionul gain control. The results of digital simulation and fact execution demonstrate that the application of ANN in CNCS is reasonable, available and effective. Moreover, the ANN control can be applied real time and in time, by the way of hardware technolngy,