Reduction of harmonics produced by electric vehicle charging stations with active power filters: An analysis and simulation study

Due to the increasing energy demand, carbon-based fuel reserves are rapidly depleting, which leads to difficulties in the energy supply chain. For this reason, developed countries encourage the production of electric vehicles (EVs). Power electronic circuits used in EVs contain nonlinear elements and these elements cause harmonic production. Harmonics cause distortion in power system processes. Filters play an important role in reducing harmonics. In this regard, active filters come to the fore due to various disadvantages of passive filters. In this thesis study, two different parallel active filter models were made to reduce harmonics caused by electric vehicle battery charging systems. An electric vehicle battery charging model with a constant current charging profile was developed. In the simulation study conducted in this model, the total harmonic distortion (THB) value was found to be at the level of 19 percent. With the proposed artificial neural network (ANN) controlled parallel active filter, this THB value was reduced to the level of 4.66 percent. In addition, in the second parallel active filter model, the synchronous reference frame method was used instead of the ANN, and the THB value was reduced to the level of 1 percent. As a result, a successful result was achieved by reducing the THB value in the simulation system with both parallel active filter models below the limit value of 5 percent determined by international organizations.