Intelligent Powertrain Control of PMSM-Based Electric Vehicles Using an Asymmetric Control Strategy

Abstract

This paper proposes a control method for electric vehicles’ powertrains that concentrates on improving the efficiency of the drivetrain as a whole by targeting regenerative energy recovery. The design of the system is based on a traditional six transistor voltage source inverter using an asymmetric control scheme together with a hybrid CNN-TD3 controller for speed control purposes. The asymmetric control method distinguishes the dynamics of the two operating modes of the inverter, that is, the traction and regenerative braking modes, by controlling the gain values based on the operating mode of the system. The TD3 algorithm adjusts the values of three continuous control parameters, Kpv, Kiv, and Ks, in real time. A multi-objective reward function aims to maximize the system’s energy efficiency, regenerative energy recovery efficiency, speed control accuracy, driving comfort, current harmonics reduction, and battery safety. Simulations performed using the WLTP Class 3 driving cycle show energy efficiency of 15.3% (25.2 to 21.44 kWh/100 km), 59.5% speed regulation error reduction, 19.6% increase in regenerative recovery efficiency from 65.2% to 77.34%, and 34.2% reduction in current THD from 8.58% to 5.65%.