In order to reduce the harmonics of the distribution system, this research examines a 3-, 3-wire ShuntActive Power Filter (SAPF) with smart controllers based on Artificial Neural Networks (ANN). A feed forward-type (trained by a back propagation method) ANN-based is built in order to enhance the performance of the traditional controller based on hysteresis controller and benefit from smart controllers. The suggested method primarily relies on the capacitor energy concept to maintain the DC link voltage of a shunt connected filter and, as a result, shortens the transient reaction time when the load abruptly changes. The proposed control method is often suitable for any other active filters of any kind. The suggested scheme's complete power system block set model was created in the MATLAB environment. Simulated actions are taken By utilising MATLAB, it is seen that the ANN driven filter reduces the%THD from 29. 71% to 2. 27%. The results of the simulated experiments also demonstrate the effectiveness of the innovative control approach in minimising harmonics, in addition to how simple it is to compute and use.