Optimal electric vehicle charging stations placement in distribution systems

This paper presents an algorithm dedicated to the electric vehicles (EVs) charging stations placement optimization in a given distribution system using genetic algorithms (GA), where daily time varying loads are considered together with random EVs charging patterns including starting time, duration, and power of charging. The problem is formulated as a non-differential combinational optimization problem, where the system losses to be minimized subject to capacity and system operation constraints. The placement alternatives considered are the installation of Level 2 single-phase slow chargers. In the GA evolutionary process, all individuals´ fitness is analyzed and for each feasible solution, a non-linear three phase power flow problem is solved and the system losses are calculated. A practical distribution system composed of 20 buses was used to validate the algorithm and demonstrate its applicability to large systems.