ANN controlled 4-leg VSC based DSTATCOM for power quality enhancement

In this paper, a novel hybrid control approach has been considered which includes an artificial neural network (ANN), a zigzag transformer and hysteresis current controllers (HCC) for a distributed static compensator (DSTATCOM). The hybrid controller is used for reduction of line current harmonics and for improvement of power factor in the system. The proposed controller pushes the source current into a sinusoidal one and tries to make it in phase with line voltage. Two main controls are offered for it: estimation of weighting factor of three phase load current and hysteresis switching algorithm for driving the voltage source converter (VSC) of DSTATCOM. The ANN with a suitable learning rate has been used to determine the compensator optimization. A zig-zag transformer is employed for neutral current reduction in the system. Simulation results are obtained for validation of the proposed DSTATCOM in the entire power system using MATLAB/Simulink and Power system tool boxes.