Robust Nonlinear Control Based on Artificial Intelligence for Electric Vehicles Under Several Constraints

Abstract

In this paper, we propose a neural network algorithm to eliminate uncertainties caused by dynamic compensators used to linearize permanent magnet synchronous motors for electric vehicle traction. To do this, we propose a speed controller based on an artificial neural network algorithm to approximate the dynamics and correlate the uncertain environmental parameters. We first estimate these uncertainties and use them as inputs to the ANN. We then find an adaptation law to eliminate the uncertainties caused by the compensators. The effectiveness and success of the proposed approach compared to classical controllers in different scenarios of electric vehicle traction are demonstrated by simulations of the adaptive ANN on Matlab/Simulink.