EV Speed Tracking Using the Regression-Based Supervised Machine Learning Algorithms

Electric vehicle (EV) technology is an emerging eco-friendly solution that has reshaped the transportation sector. This paper aims to design a controller for EV speed tracking using three regression-based supervised machine learning (ML) algorithms. Regression techniques such as Ensemble Learning (EL), Support Vector Regression (SVR), and Gaussian Process Regression (GPR) are used to develop ML-based controllers to predict the EV speed. The training data set for the ML models, including the predictors and response data, is extracted from the fuzzy-based controller scheme. Statistical measurement metrics, including the correlation coefficient (R), Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Square Error (RMSE), are used to evaluate the performance of the models. The simulation of the proposed controllers is carried out in a MATLAB-Simulink environment. Simulation results demonstrate the GPR model's advantage over other ML models for tracking EV speed. From the statistical performance of the EL, GPR, and SVR models, it is observed that GPR model with Bayesian optimization gives the best performance with the lowest RMSE of 3.31 and MAE of 0.0026 compared to other models.