高速直线感应电机自学习抗扰控制策略  

Self-Learning Anti-Disturbance Control Strategy for High-Speed Linear Induction Motor

作　　者：徐飞[1,2,3] 姜新宇 李子欣 史黎明[1,2,3] 李耀华 Xu Fei;Jiang Xinyu;Li Zixin;Shi Liming;Li Yaohua(State Key Laboratory of High Density Electromagnetic Power and Systems Institute of Electrical Engineering Chinese Academy of Sciences,Beijing 100190 China;University of Chinese Academy of Sciences,Beijing 100049 China;Institute of Electrical Engineering and Advanced Electromagnetic Drive Technology QILU ZHONGKE,Jinan 250100 China)

机构地区：[1]高密度电磁动力与系统全国重点实验室中国科学院电工研究所,北京100190 [2]中国科学院大学,北京100049 [3]齐鲁中科电工先进电磁驱动技术研究院,济南250100

出　　处：《电工技术学报》2025年第6期1771-1783,共13页Transactions of China Electrotechnical Society

基　　金：中国科学院稳定支持基础研究领域青年团队计划资助项目(YSBR-045)。

摘　　要：高速直线感应电机在跨声速运行时,激波非定常特性引发的抖振现象易导致电机动子速度测量受到强烈不确定扰动,推力控制失稳、跨声速运行失败。该文首先建立了高速直线感应电机自学习数学模型,由历史实验数据学习电机模型参数,依据数学模型对直线感应电机动子速度进行高精度多步预测。其次由预测速度与传感器量测速度构建滚动预测矩阵,提出基于矩阵列向量变异程度的量测速度可信度计算方法。最后依据预测速度、量测数据及可信度值加权运算,提出不确定扰动下高速直线感应电机自学习抗扰控制策略。仿真结果验证了自学习抗扰控制策略对短时不确定扰动有较好的抑制效果,在跨声速时修正速度与实际速度误差小于0.7%,有效解决了不确定扰动导致的电流分量偏离和推力失稳问题,原理样机实验结果验证了控制策略对不确定扰动抑制的有效性。High-speed linear induction motors have the advantages of light rotor mass,simple structure,and high reliability,which are suitable for ground ultra-high-speed test facilities and electromagnetic launch fields.However,when high-speed linear induction motors run at transonic speeds,the unsteady aerodynamic characteristics of shock waves will cause strong vibrations in the rotor,resulting in strong uncertain disturbances in the position and speed measurement of the linear motor rotor and the electromagnetic mechanism.This uncertain disturbance is short-term,highly dynamic,sudden,and unpredictable.Consequently,slip control of the linear induction motor is abnormal,thrust control is unstable,and the transonic operation fails,seriously restricting the system speed improvement.This paper establishes a self-learning mathematical model of a high-speed linear induction motor.The parameters of the motor model are learned through historical experimental data to realize high-precision multi-step prediction of the mover speed.The prediction speed constructs the rolling prediction matrix,the sensor measurement speed is calculated,and a method for reliability evaluation of the measurement speed is proposed based on the variation degree of the matrix column vector.Finally,according to the weighting of the prediction velocity,the measurement data,and the credibility value,a self-learning anti-disturbance control strategy for a high-speed linear induction motor under uncertain disturbance is proposed to realize the anti-disturbance control under transonic speed conditions.The simulation and experimental results show that the proposed model can realize linear motor rotors’high-precision multi-step rapid speed prediction.The speed correction method blocks the propagation of uncertain disturbances and accurately suppresses uncertain disturbances.The rotor speed’s simulation error is less than 0.7%,and the experimental error is less than 1%.Under the uncertain disturbance of e=0.2,the proposed self-learning anti-disturbance contro

关 键 词：不确定扰动 直线感应电机 抗扰控制 自学习模型 

分 类 号：TM346[电气工程—电机]