Kalman Filter based Neural Network Control Strategy with PI Controller for Optimum Performance of Series Active Power Filter

Harmonics and reactive power have been a key disquiet in power systems owing to the copious usage of power electronic converters during varied levels of its working. This increase will suppress the quality of the power of the system and there will be contamination in the power system leading to lesser efficiency and throughput. An active power filter when connected in series to the power system will help out in diminishing the harmonics within the system and appeasing the matter of reactive power throughout. Here, in the paper a neural network controller which is based on Kalman filter is connected to the active power filter to reduce the Total Harmonic Distortion [THD] to 3.61 % and after that again a PI controller is added to it to observe that betterment is achieved in THD again to about 1.11 %.