基于平行度误差补偿的多通道双晶单色器运动控制

Motion control of multi-channel double crystal monochromator based on parallelism error compensation

针对新型多通道双晶单色器系统易受晶体片平行度误差影响导致性能失效的问题,提出一种迭代学习模型预测控制器(Iterative Learning Model Predictive Control,ILMPC)策略。该策略将主-从轴晶体片平行度误差转换为从轴电机的重复参考运动轨迹,采用ILC迭代消除周期性机械安装误差,再利用MPC对单个转动周期内从轴电机的随机性误差进行抑制。最终在单轴电机运动实验平台上对该策略进行轨迹跟踪验证。实验结果表明,在晶体片有效工作区间内从轴电机的位置跟踪误差达到1.44″,相比于PID控制、MPC、DOB+MPC以及ILC+PID组合控制器策略,分别降低了99.64%,98.52%,98.26%和73.33%。研究结果验证了所提出的ILMPC策略在补偿晶体片平行度误差和提高平行对准精度方面的有效性,该策略对于改善新型多通道双晶单色器性能具有较好的实际应用价值。

This paper proposes a control strategy to address the issue of crystal parallelism errors that affect the performance of the new multi-channel double-crystal monochromator(DCM)system.The control strategy integrates iterative learning control(ILC)and model predictive control(MPC)to reduce random position errors of the slave motor and eliminate repetitive mechanical installation errors,respectively.The parallelism errors between the master and slave crystals are converted into the repetitive reference motion trajectory of the slave motor,and the MPC is used to improve the slave motor position tracking within a single rotation cycle.Meanwhile,ILC iterations are applied during rotation cycles to eliminate repetitive mechanical installation errors.The proposed control strategy is verified using a single-axis motor motion experimental platform.Experimental results demonstrate that the motor position tracking error reaches 1.44″–reduction by 99.64%,98.52%,98.26%,and 73.33%as compared to the errors of PID,MPC,DOB+MPC,and ILC+PID combination controller strategies,respectively.The proposed control strategy effectively compensates for crystal parallelism errors and improves parallel alignment accuracy, providing practical application value for the performance enhancement of the new multi-channel DCM system.