【摘要】 由于电机本体、驱动电路及机械连接等多环节因素影响，传统PI(proportional integral)控制下永磁同步电机在低速运行时易出现较大的转速波动。首先，建立多源性力矩扰动模型与分析转速波动机理，将扰动分为与机械角相关的周期性干扰与非周期低频干扰两部分；然后，提出一种结合迭代学习控制与扰动观测器的扰动抑制策略，利用控制误差对周期性扰动进行在线学习，利用反馈状态量结合学习信息对低频非周期扰动进行观测，将两部分结果共同补偿到电磁力矩，实现对多源性扰动的抑制；最后，在给出所提控制器结构基础上，分析了所提控制策略的有效性与收敛性，给出控制器主要参数设计方式。试验结果表明，所提控制策略相较于传统控制策略整体动态性能相近，在稳态时相较于PI(proportional integral)控制将转速波动抑制到1/4以下，相较于传统抑振策略整体转速波动在1/2以下，在负载力矩突变时相较于PI控制速度变化在1/2以下，证明所提算法在多源性扰动抑制方面的有效性与优越性。更多还原

【Abstract】 Due to effects of multiple factors of motor body, drive circuit and mechanical connections, permanent magnet synchronous motor under traditional proportional integral(PI) control is prone to larger speed fluctuation during low-rotating speed operation. Here, firstly, a multi-source torque disturbance model was established and mechanism of speed fluctuation was analyzed, disturbance was divided into two parts of periodic interference related to mechanical angle and non-periodic low-frequency interference. Then, a disturbance suppression strategy combining iterative learning control and disturbance observer was proposed. Periodic disturbance was learned online using control error, and low-frequency non-periodic disturbance was observed using feedback state variables combined with learning information. The two results were jointly compensated to electromagnetic torque to realize multi-source disturbance suppression. Based on giving the structure of the proposed controller, the effectiveness and convergence of the proposed control strategy were analyzed, and the design method for main parameters of the controller was given. The experimental results showed that the proposed control strategy has similar overall dynamic performance compared to traditional control strategies; in steady state, speed fluctuation is suppressed to less than 1/4 compared to PI control; compared to traditional vibration suppression strategies, the overall speed fluctuation is suppressed to less than 1/2; when load torque suddenly varies, speed fluctuation is suppressed to less than 1/2 compared to PI control; the effectiveness and superiority of the proposed algorithm in suppressing multi-source disturbance are verified.更多还原