Model Predictive Control of Demagnetization Vibration in Permanent Magnet Synchronous Motor

Abstract

This article proposes a new nonlinear processing method combined with model predictive control (MPC) to study the demagnetization vibration problem of permanent magnet synchronous motors (PMSMs). First, based on the theory of rotor dynamics, a demagnetized rotor vibration dynamics model considering the unbalanced magnetic pulling (UMP) caused by demagnetization of PMSM was constructed. Accurately modeling the nonlinear UMP function is difficult in practical engineering, thereby increasing the complexity of accurately constructing the parameters of the demagnetized rotor vibration dynamics model. Thus, this article innovatively constructs an empirical function model for UMP based on practical experience and then establishes a system observer model. Sufficient conditions are obtained for the system observer model based on Lyapunov stability theory to effectively observe the demagnetization vibration dynamics model. Then, an MPC scheme is proposed based on the constructed demagnetization vibration observer model. Finally, numerical simulation was conducted to verify the effectiveness of the system observer model in tracking the system model and predicting the effectiveness of system vibration control.