Experimental Investigation on Electric Vehicle Braking System Using Machine Learning for Enhanced Performance and Safety

The rapid increase in the popularity of Electric Vehicles (EVs) can be attributed to their environmental benefits and innovations. Regenerative braking is a braking method considered a very important safety feature in EVs. This study presents an empirical evaluation of an EV braking system, where close attention is paid to the implementation of machine learning (ML) methods to maximize operational efficiency and enhance safety. Practical driving tests were conducted, and sensor data of various parameters, such as vehicle speed, brake pedal pressure, motor torque, and battery charge, were recorded. Various machine learning algorithms were tested, including Artificial Neural Networks (ANN), Support Vector Regression (SVR), and Random Forests, for predicting braking distance and optimizing the combination of regenerative and friction braking. The results indicate the immense potential of machine learning to maximize braking efficiency, reduce wear and tear on friction brakes, and improve overall safety in EVs.