This paper studies a new energy vehicle-pile matching method based on privacy protection technology, aiming to solve the user privacy leakage problem encountered in the charging process of new energy vehicles. In order to protect user data privacy, a privacy protection method based on OT-Extension bidirectional authentication protocol, private set intersection (PSI) and private information retrieval (PIR) cryptography techniques is proposed to realize effective interaction and query of pile data. Based on XGBoost model, a car pile matching model is constructed using federated optimization to optimize the matching of charging piles. By training and predicting the data of user's charging demand and charging pile state, the car pile matching model can match charging pile efficiently. Finally, security analysis and performance evaluation are carried out on the results of the proposed pile matching model, and comparison with logistic regression model proves the reliability, feasibility and effectiveness of the privacy protection technology in this paper in the pile matching method.