Wavelet-Enhanced Active Disturbance Rejection Control for PMSM in Low-Speed Textile Processing

This paper addresses the speed fluctuation problem of permanent magnet synchronous machines (PMSMs) under low-speed conditions in high-precision fabric and leather processing. An improved Active Disturbance Rejection Control (ADRC) algorithm is proposed. Based on a tracking differentiator-extended state observer-nonlinear state error feedback law structure, a wavelet packet analysis-enhanced adaptive extended state observer (AESO) is used to estimate and compensate for composite disturbances. A hyperbolic tangent nonlinear feedback law is applied to suppress chattering, and a recursive least squares (RLS) algorithm is employed for online identification and feedforward compensation of time-varying parameters. The integral absolute error (IAE) is reduced to 0.019 rad��s, and the ability to suppress disturbances below 5 Hz is significantly improved. The system maintains stable control under a +50% large parameter disturbance. This method improves the operational stability of low-speed textile and leather processing tasks, achieving higher fabric quality, reducing material waste, and improving the preci-sion of automated manufacturing.