Performance Investigations of Adaptive ANN-SVPWM Controlled Shunt Active Power Filter

Efficient and Adaptive Artificial Neural Network (ANN) based Shunt Active Power Filter (SAPF) for mitigating harmonics is presented in the paper. ANN based reference current generation is implemented using Neural Synchronous method while Levenberg Marquardt method is used for training algorithm. ANN based Space Vector Pulse Width Modulation (ANN-SVPWM) controller is realized for the effective control of SAPF. Performance of the proposed ANN-SVPWM controller is compared with conventional Hysteresis Current Controller (HCC). Hence, both the reference compensating currents as well as controller of the proposed SAPF are realized using Artificial Neural Network. Effective compensation of the load harmonics and good dynamic response of the ANN based SAPF is proved by the presented simulation results using MATLAB. The proposed SAPF effectively compensates for the load harmonics in steady-state as well as during dynamic conditions.