Palliation of Current Harmonics by Employing Optimised FO (PI+PD) Cascade Controller Regulated SAPF with the Exertion of Taguchi SNR Method.

Abstract

In present world, application of non-linear devices is increasing at a very rapid pace. These non-linear devices actually tend to draw currents that are non-sinusoidal in nature. As a result of this harmonic distortions occur in electrical networks. There are various techniques that are used in order to remove harmonics from current, improve power factor and compensate reactive power. Shunt active power filter (SAPF) is very renowned amongst those techniques. However, SAPF is still one of the thrust areas of study because designing an extremely effective SAPF is highly problematic. Choosing correct current control strategy, right technique for the production of switching pulses and best set of parameters can be difficult jobs to do while designing SAPF. We had researched a lot about the impacts of important designing parameters like capacitance of DC-link, SAPF impedances and different constraints of controller. We used Taguchi SNR method to get optimized values of constraints of controller. The advantages of using the Taguchi technique for optimizing SAPF variables over existing adaptive network based fuzzy inference systems (ANFIS) are discussed too in this research paper. Instead of using conventional PI controller, we employed fractional order (PI+PD) controller. The PQ theory was applied to extract the reference currents. Whereas, the gate pulses that are responsible for controlling switches of voltage source inverter were attained using hysteresis current control strategy. As a result, the full SAPF model was created and simulated in MATLAB/Simulink. Both the anticipated Taguchi SNR method and already existing ANFIS technique were applied in order to optimize the constraints of SAPF. And finally, by taking the optimized values of constraints, results were captured, examined and the supremacy of the anticipated Taguchi method is discussed here.