線型馬達之高速伺服控制

Abstract

本文針對線型步進馬達的模型、控制器硬體及控制的手段加以改進，使用精度為 1 μm 的光學尺，可以提高線型馬達的操作速度到 1 m/s，零速時的位置命令誤差在 ±1 μm 。首先是馬達模型加入適當的阻尼力及四階的齒槽效應力，因此使得馬達數學模型的模擬結果和實驗的結果相符合。由於更換介面卡的類比數位轉換器等等的改變，可將取樣頻率大幅提高。在控制器的設計方面，本文主要以全階模型的雙迴路控制法與降階模型控制法為主。使用全階模型所設計的控制器，必須補償電壓向量的角度，才能提高操作速度。而使用降階模型控制器，所必須補償的電壓向量角度較小，比全階模型控制器更容易使線型步進馬達操作在高速下。另一方面，因為適應控制法具有參數估測的能力，所以用來做馬達的參數鑑定。

This thesis studies the high-speed servo control of linear pulse motors. To achieve this objective, a more accurate mathematical model of the motor with a damping term and up to fourth-order periodical cogging forces is identified and used in the design of the controller. As for the control algorithm, we have implemented both a two-loop control method and a reduced-order control method. The angle of the voltage vector is also increased corresponding to the speed of the motor to compensate the delay in the sampling and control process. Finally, the sample rate of the closed-loop control system is further increases up to 26KHz by changing the AD converter of the control interface board. With the above improvements in the control system, we have demonstrated in both simulation and experimentation of a trajectory tracking task that, the highest speed of the linear pulse motor can be increased to 1 m/s while the position error can be reduced to ±1 μm at zero speed.