Charging Data Analysis and Clasification for Microgrids

Abstract

This thesis aims to develop a classification model for electric vehicles (EVs) based on data from EV charging stations.The study utilizes a dataset of 6 charging sessions from EV charging station and implements two deep learning algorithm including LSTM and Auto-Encoders to classify EVs. The performance of the classification model is evaluated based on accuracy rates & precision.The study also identifies key charging characteristics that are most significant in distinguishing between different types of EVs, including charging time, energy consumption, and charging patterns.The findings of this research have significant implications for the development of EV charging infras- tructure and services. The classification model developed in this thesis can be used to optimize charging station operations, improve charging services, and develop EV adoption strategies. The study also highlights the importance of utilizing data from EV charging stations in understanding the EV market and improving the efficiency of charging infrastructure.