Wavelet Neural Network Control of Synchronous Motor Systems in an Electric Vehicle

The growing adoption of electric vehicles as a sustainable alternative to fossil-fuel-based transportation has intensified the demand for high-performance motion control solutions for their propulsion systems. Synchronous motors coupled to mechanical power transmission systems constitute the core electromechanical powertrain in a wide range of electric vehicle configurations. In this paper, a novel wavelet neural network control framework for robust tracking of reference trajectories of angular velocity generated online, electromagnetic torque and electric currents in a permanent magnet synchronous motor used as a motion actuator in an uncertain electric vehicle drivetrain is introduced. Wavelet neural networks are incorporated as an alternative adaptive mechanism to tune the motor control bandwidth parameter online. In this fashion, automatic adjustment of the control parameters can be performed in response to variable operating conditions and uncharacterized external vibration disturbances affecting uncertain electric vehicle dynamics. The effectiveness of the presented control framework based on wavelet neural networks for synchronous motors under nonstationary dynamic torque disturbances induced by a power transmission system of an electric vehicle is validated by numerical results. Several case studies for simultaneous tracking of dynamic reference trajectories of angular velocity, electromagnetic torque and electric currents generated online for various uncertain variable operational scenarios of an electric vehicle are developed.