Adaptive Linear Predictive Deadbeat Control Against Machine Parameter Uncertainty of SPMSM Drives for Electric Vehicles

Abstract

A fast and accurate torque/current response is important for optimizing the efficiency of electric vehicles (EVs) and promptly implementing braking instructions in emergencies. Deadbeat predictive control has attracted much attention in electrical drives for EVs due to its excellent dynamic performance. However, parameter mismatch significantly influences the performance of conventional deadbeat predictive controllers. Existing solutions can effectively suppress the steady-state error caused by parameter mismatch, but their abilities to improve transient performance against parameter uncertainty are limited. To address this issue, an adaptive linear predictive current deadbeat controller for surface-mounted permanent magnet synchronous motor (SPMSM) drives is proposed in this article. The actual current response characteristic is first tested by applying a test voltage vector. Thereafter, the final required voltage command, which can bring the current to its new reference, is determined by utilizing the actual machine current response characteristic derived from the test voltage vector. The proposed method is simple to implement and can be combined with many advanced methods to achieve both satisfactory dynamic and steady-state performances against parameter uncertainty. The effectiveness and compatibility of the proposed method have been verified by comparing it with other two advanced deadbeat predictive control methods under different dynamic and parameter mismatch conditions.