Enhanced Performance of Electric Vehicle Powertrains Using Hybrid Neural Network Models

As oil reserves decline, fossil fuel transportation's political, economic, and environmental issues worsen. EVs, a new green technology, might reduce oil use, increase sustainable road transport, and reduce greenhouse gas emissions. Fault categorization is essential for accelerating electric vehicle power train installation by improving efficiency and reliability. To solve fault classification problems, significant data preparation was used to prepare the dataset for model training. These algorithms removed missing values, found duplicates, and transformed data. Cross-validation was used to train and evaluate different neural network (NN) algorithms, including BiRNN, CNN-GRU, GRU, BiGRU, and BiCRNN, to assess performance. BiCRNN has the highest accuracy of the models we examined, 97.95%. This confirms its EV power train fault detection superiority. The model's impressive performance suggests it may be used to ensure electric vehicle efficiency and reliability. This paper emphasizes the value of advanced neural network models, particularly BiCRNN, in EV power train fault classification. The findings emphasize the importance of adopting cutting-edge machine learning approaches to promote greener transport systems to address important energy and environmental challenges.