基于前馈及反馈补偿的高性能直线伺服系统

Study on Feedforward and Feedback Compensation for High Performance Linear Servo System

数控机床用直线电动机在工作过程中,推力波动、负载扰动和非线性摩擦等各种扰动因素降低了直线伺服的性能。为抑制扰动满足数控机床直线伺服系统速度跟踪性能的要求,在分析永磁直线同步电动机数学模型和各种扰动因素的基础上,引入具有动态递归特性的Elman神经网络,设计出具有前馈及反馈综合补偿效果的复合控制器,不仅可以对各种扰动进行反馈补偿,而且对速度给定信号进行预测前馈补偿控制,提高了直线伺服系统的跟踪精度和响应速度,仿真结果验证了Elman神经网络的动态性能及综合补偿的优点,表明了复合控制方案的有效性和可行性。

Disturbance variations,such as force ripple,load disturbance and nonlinear friction,directly impose on the motor shaft,leading to significant effect on the computer numerical control （CNC） machine tool servo system performance.In order to satisfy the requirements for speed command tracking performance and eliminate the influence of disturbance,on the bases of the analyzing the mathematics model of permanent magnet linear synchronous motor （PMLSM） and disturbance variation,Elman dynamic recurrent Neural Network was introduced and compound controller with feedforward and feedback comprehensive compensation was proposed.The disturbances were compensated by feedback component and reference speed input signal was preview controlled by feedforward component.The tracking performance and respect speed of linear server system were improved.The advantages of the comprehensive compensation and dynamic of Elman NN were proved.The simulation results show that this compound control scheme is feasible and effective.