[关键词]
[摘要]
与传统单层气隙电机相比,双定子磁场调制(FMDS)电机存在较大的转矩脉动。转矩脉动随转子位置的变化而周期性变化,从而导致速度脉动。针对周期性转矩脉动、快速性与超调性的矛盾以及外部不确定干扰的存在,提出一种基于自抗扰迭代学习(ILC-ADRC)的转速外环控制器。通过设定期望转速的过渡过程,避免了阶跃输入引起过大转速超调,降低了迭代学习的初始条件要求。设计线性扩张状态观测器(LESO)以简化参数调节。加入过去周期性的转速误差信息,以补偿转矩脉动。最后,利用状态误差反馈控制律生成控制信号。仿真结果表明,该系统响应快速、无超调,有效地降低了电机的转矩脉动。
[Key word]
[Abstract]
Compared with the traditional single layer air-gap motors, larger parasitic torque ripple exists in field-modulation double-stator (FMDS) motor. The torque ripple varies periodically with rotor position, which leads to speed ripple. For the existence of periodic torque ripple, the contradiction between response and overshoot, and the presence of external uncertain disturbances, a novel outer-loop speed controller based on iterative learning control and active disturbance rejection control (ILC-ADRC) is proposed. By arranging the transition process of expected speed, the excessive overshoot of speed caused by stepped input is avoided, and the initial condition requirements of iterative learning are reduced. A linear expanded state observer (LESO) is designed to simplify parameter adjustment. The past periodic rotational speed error information is added to compensate the torque ripple. Finally, the state error feedback controller is used to generate control signals. The simulation results show that the system responds quickly without overshoot, and the torque ripple of the motor is effectively reduced.
[中图分类号]
[基金项目]
国家自然科学基金项目(51777216)