Optimal Planning and Operation of Electric Vehicles Battery Swapping Stations

Abstract

Electric Vehicles (EVs) nowadays have become increasingly prevalent due to the advancements in EV technology and their impact on reducing greenhouse emissions. However, there are still some factors affecting the fast deployment of EVs such as the limited driving range and the charging time. Due to the limited driving range, EVs need to be charged frequently, but charging requires a long period at traditional EV charging stations, whereas fast-charging stations still have concerns regarding the wait and the charging time, which might cause traffic jams near the station. In this thesis, new dynamic optimal operation and planning approaches of EV battery-swapping stations (BSS) are introduced. In the operation phase, the goal is to maximize the daily profit using a rolling horizon optimization (RHO) mechanism and determining the optimal operating schedule for swapping and charging/discharging processes. The problem is formulated as mixed-integer linear programming (MILP) problem with nonlinear battery degradation characteristics included. Long-short-term memory (LSTM) recurrent neural network is used as a time series forecasting engine for predicting the EVs' arrivals. The proposed approach is tested and compared with the unscheduled operation and day-ahead scheduling. The results show that the dynamic operations scheduling using the proposed RHO mechanism results in a higher profit. In the second phase, an optimal planning approach for a photovoltaic-based BSS system is proposed considering the PV system and EV arrivals uncertainty. The main goal of the planning part is to determine the optimal size of the BSS assets and to optimally allocate the BSS in the distribution network. Markov Chain Monte Carlo Simulation is used to tackle the uncertainty associated with photovoltaic output and EV arrivals. Simulation results show the effectiveness of the proposed BSS system and an optimal solution is obtained which maximizes the annualized profit.