Two-layer optimal charging strategy for electric vehicles in old residential areas

Electric vehicle (EV) departure time is an important variable in current coordinated charging studies. The predicted value of EV departure time is more reliable than the user-set departure time. In this study, a long short-term memory prediction model is used to accurately predict the departure time, and a two-layer optimal charging strategy is proposed. Total user satisfaction is set as the objective function, with the constraint of a safe distribution network operation. The proposed strategy is tested using a set of real EV travel data in an old residential area, and its performance is comprehensively compared with two alternative charging strategies, namely, uncontrolled charging and two-layer optimal charging with a set departure time. The proposed strategy outperforms the rival strategies by improving total user satisfaction, while ensuring safe distribution network operation in the old residential areas.

Automated summary

This study uses a long short-term memory prediction model to improve the accuracy of electric vehicle departure time prediction and proposes a two-layer optimal charging strategy. Tested with real EV travel data in an old residential area, the strategy outperforms rival strategies by increasing user satisfaction while ensuring safe distribution network operation.