Smart Charging of Electric Vehicles on the Grid Based on Social Spider Swarm Intelligence Algorithm

This paper investigates the impact of large-scale, uncontrolled charging of electric vehicles (EVs) on the grid distribution network. This is a growing concern for the safety and reliability of the grid. To address this problem, first, we conducted a quantitative data analysis using real-time, large-scale datasets of EVs charging sessions from public charging stations to understand the impact of EVs charging on the grid. Secondly, we exploit the flexibility potentials to minimize the uncontrolled charging of EVs on the grid by time-shifting EVs charging. Third, a framework based on the social spider optimization (SSO) algorithm was adapted to improve EVs search process for an optimal power supply and the best time-of-use application that will minimize EVs energy consumption. Fourth, as an application example, we used statistical model of time series analysis to evaluate the implementation of the algorithm and compared results for two case studies. One case study with uncontrolled charging for a 24hr period showed a cumulative energy consumption by EVs to be 409.97kWh. The second case study based on a controlled charging with two different scenarios. Scenario-1 was based on time-shifting EVs charging to evening hours and after peak periods. The results shows a 47.7% reduction in total energy usage and cumulative energy consumption by EVs to be 195.57kWh when compared with the case of uncontrolled charging. Scenario-2 was based on shifting EVs charging from the utility grid to solar PV microgrid 9am to 5pm. The results shows a 55.7% reduction in the total energy usage and a cumulative energy consumption of 228.37kWh when compared to the case of uncontrolled charging. This study presents a brief description of an ongoing research.