Optimization of electric vehicle charging times in the distribution network with genetic algorithm

Because of their environmental friendliness, reducing dependence on fossil fuels and advantage of fuel economy, Electric vehicles are supporting by many countries and getting more interest of vehicle users and vehicle manufacturers every day. Although their number is small for now, the ever-increasing electric vehicles, which are predicted to cover 58% of the total number of vehicles in traffic in 2040, bring along some problems as well as their advantages. Connection of electric vehicle charging stations to existing distribution networks; will cause adverse effects such as voltage drop, frequency fluctuation and network black out. In this study, Genetic Algorithm, which requires only the objective function and works according to probability rules, is used as an optimization method for solving complex problems involving multiple variables such as capacity values of electrical networks and hourly changes of grid load curves. Genetic Algorithms is an optimization method based on natural selection principles. It performs an efficient search by scanning a certain part, not the entire solution space, and reaches a solution in a shorter time. In this thesis, optimum charging stations load profiles were generated with the Genetic Algorithm and compared with uncoordinated load profiles on the test feeder.