【目的】针对永磁同步直线电机在多源复合扰动下存在的动态响应慢、跟踪精度不足及参数整定复杂等问题,提出一种基于粒子群优化算法的改进降全阶切换自抗扰控制(PSO-RSADRC),旨在突破快速整定时间与抗干扰性。【方法】首先,构建智能切换自抗扰控制(SADRC)架构,通过线性/非线性动态切换优化扰动抑制性能。然后,提出降全阶切换自抗扰控制(RSADRC),简化观测器结构以降低参数耦合度。最后,为解决非线性误差反馈模块的多参数耦合问题,开发PSO-RSADRC,并基于四阶S型运动规划模型开展仿真与试验验证。【结果】仿真结果表明,PSO-RSADRC在1 A阶跃扰动下的时间加权绝对误差积分降至0.95,最大动态误差仅为19 μm,较传统线性自抗扰控制精度提升98%。试验平台验证中,整定时间缩短至57 ms,响应速度提升32%。【结论】所提出的PSO-RSADRC策略有效解决了多源扰动下的精密运动控制难题,其参数整定效率与抗扰鲁棒性显著优于传统方法,为微电子封装等高精度工业场景提供了可靠的技术方案。

Abstract: [Objective] This study addresses the challenges of slow dynamic response, insufficient tracking accuracy, and complex parameter tuning in permanent magnet synchronous linear motor under multi-source disturbances. Reduced-full-order switched active disturbance rejection control based on particle swarm optimization (PSO-RSADRC) algorithm is proposed to enhance control precision. [Methods] First, an intelligent switching active disturbance rejection control (SADRC) architecture was constructed, optimizing disturbance rejection performance through linear/nonlinear dynamic switching. Then, reduced-full-order SADRC (RSADRC) to simplify the observer structure and reduce parameter coupling was proposed. Finally, to address the multi-parameter coupling issue in the nonlinear error feedback module, a PSO-RSADRC approach was developed, and simulation and experimental validation were conducted based on a fourth-order S-curve motion planning model. [Results] Simulation results demonstrated that PSO-RSADRC achieved an integral time-weighted absolute error of 0.95 and maximum dynamic error of 19 μm under 1 A step disturbance, improving precision by 98% compared to conventional linear ADRC. During platform validation, the settling time was reduced to 57 ms with a 32% enhancement in response speed. [Conclusion] The proposed PSO-RSADRC strategy effectively resolves precision motion control challenges under multi-source disturbances, demonstrating significantly superior parameter tuning efficiency and disturbance rejection robustness compared to conventional methods. This solution provides a reliable technical approach for high-precision industrial applications such as microelectronics packaging.

国家自然科学基金(61973093)