摘要
为解决永磁同步电机无位置传感控制中传统滑模观测器易产生高频抖振与估算精度不高的问题,设计了一种基于改进型超螺旋算法的连续可导滑模观测器,并引入锁相环实现高精度的位置信号提取。该方法通过引入全程连续可导且具备饱和特性的切换函数,替代传统不连续符号函数,显著削弱了系统中的抖振现象,同时增强了系统对扰动的抑制能力。在MATLAB/Simulink环境下搭建完整控制仿真平台,对比双曲正切函数滑模结构与所提方法,结果表明:所提方法在转速跟踪响应中几乎无超调,扰动下恢复速度快,稳态误差显著减小;在位置估算方面误差峰值下降超过60%,曲线更加平滑贴合真实值;转矩波动幅度降低近一半,电流响应保持较高对称性与稳定性。上述结果验证了所提方法在动态性能、观测精度与系统鲁棒性方面较传统控制结构具有显著优势,为永磁同步电机在复杂工况下的高性能无位置传感控制提供了可靠的技术路径。
To address the issues of high-frequency chattering and low estimation accuracy in conventional sliding mode observers(SMOs)used for sensorless control of permanent magnet synchronous motors(PMSMs),this paper proposes a continuously differentiable super-twisting observer enhanced with a phase-locked loop(PLL)for precise rotor position extraction.The approach replaces the traditional discontinuous sign function with a smooth switching function that is both continuously differentiable and saturated,effectively suppressing chattering while improving robustness to external disturbances.A comprehensive control model is developed in MATLAB/Simulink to evaluate the proposed method against the conventional SMO based on hyperbolic tangent functions.Simulation results demonstrate that the improved observer exhibits near-zero overshoot in speed tracking,faster recovery under load disturbances,and significantly reduced steady-state errors.Rotor position estimation error peaks are reduced by over 60%,and the observed trajectory smoothly and closely matches the actual position.Torque ripple is nearly halved,and three-phase current profiles maintain high symmetry and stability.These findings confirm that the proposed method offers notable advantages in dynamic response,estimation precision,and system robustness,making it a promising solution for sensorless PMSM control in complex operating conditions.
作者
刘旺
王春梅
李秀
李佳鑫
LIU Wang;WANG Chunmei;LI Xiu;LI Jiaxin(School of Mechanical and Electrical Engineering,Guizhou Normal University,Guiyang 550025,China)
出处
《机械与电子》
2025年第12期52-58,67,共8页
Machinery & Electronics
关键词
永磁同步电机
无位置传感控制
滑模观测器
超螺旋算法
锁相环
连续可导函数
permanent magnet synchronous motor
sensorless control
sliding mode observer
super-twisting algorithm
phase-locked loop
continuously differentiable function
