Electric Vehicles Charging System for Fast and Safe Charging using LSTM based Gradient Boosted Regression Tree

In this study, a process-based system for designing and managing charging procedures for electric vehicles (EVs) is proposed. Electric Vehicles charging should be done in a balanced manner, considering prior experience, meteorological information based on data mining, and modeling methodologies, due to the limitations of the electrical power distribution network and the lack of smart meter devices. It was also decided to develop a mobile app to aid the EV driver in information exchange and user mobility. Methods for predicting battery charge patterns through data preprocessing and regression model training. Finally, the model is trained using LSTM-GBRT, then the cuckoo search feature selection approach is applied to improve accuracy. The proposed method achieves better results than the LSTM and GRU models.