Fault Diagnosis of Electric Vehicle’s Battery by Deploying Neural Network

Abstract

Developed nations have focused more on environmental degradation and climate change in response to rising concerns about meeting the needs of their citizens. The market for emission-free Electric Vehicles (EVs) is now a key area of international rivalry and progress. Rising concerns over high voltage hazards in EVs are a direct result of their increasing popularity. It is crucial to examine the problem diagnosis method of lithium-ion batteries (LIB) because the battery system is responsible for more than 30% of EV accidents. EV’s LIB has complicated fault types that are difficult to treat. Timely and efficient battery pack problem diagnosis is crucial for ensuring the real-time safety of EV function. With the help of neural network models like Multilayer Perceptron (MLP) and Radial Basis Function (RBF), this research demonstrates a technique for detecting and fixing EV battery problems. MATLAB is used to simulate the battery and generate the necessary data for the battery failure detection system. Accuracy is improved through pre-processing the data after it has been generated. Both models are trained and then put through tests to determine how well the models are performing. By contrasting the positive and negative metrics, the best model can be determined.