Yuanhang Cao

Deadbeat predictive current control (DPCC) method is a predictive control (PC) method with excellent dynamic-state response. However, it suffers from poor steady-state performance due to parameter mismatch, unmodeled parts of the system and external disturbances. Unlike the methods utilizing model free and observer, a simple DPCC with parameter robustness improvement (SDPCC-PRI) method for permanent magnet synchronous motor (PMSM) is proposed to enhance the parameter robustness in this paper. Firstly, a simple error compensation (SEC) algorithm based on minimization of cost function designed in this paper is capable of calculating and compensating the error terms in the extended voltage prediction model in real time, thus suppressing the adverse effects caused by parameter mismatch. Furthermore, the online inductance extraction (OIE) algorithm addresses the impact of gain variation resulting from inductance mismatch on SEC algorithm. Finally, the simulation based on Matlab/Simulink and experiments conducted on 1.25-kW experimental platform are provided to validate the proposed SDPCC-PRI method.

With the aim of improving parameter robustness in the model predictive current control (MPCC) for permanent magnet synchronous motor (PMSM) drive system, a simple motor-parameter-free model predictive current control (MPF-MPCC) methodology is proposed in this study. First, for different characteristics of dq-axes prediction model, the parameters of resistance and flux-linkage in the model are removed. Besides, an estimation factor is designed to substitute the inductance term in the model and is calculated based on a simple rhythm in real time. Utilizing the established predictive current model devoid of any motor parameters, the optimal voltage vector is accurately selected for controlling the motor. Finally, the effectiveness of the proposed MPF-MPCC methodology is substantiated through rigorous validation on a 1.25-kW experimental platform.