Hybrid Active Power Filter Topologies for Power Quality Improvement

Abstract

Sustainable development of the society is measured with the help of various newlineparameters such as economy growth, social quality matrices, happiness and many other newlineparameters related to consumer satisfaction. With the advent of smart grid evolution, newlineconsumer centric policies are in spotlight. This is the main reason of restructuring of the newlinepower system also. The focus of the design engineers is to make system reliable and newlinehassle free. newlineDue to increase in system demand, along with the application of nonlinear newlinedevices, heavy burden on the grid is experienced. These effects deteriorate quality of newlinepower received at consumer end. This affirms the presence of harmonics in system s newlinefundamental voltage and currents. For mitigation of these harmonics, many technologies newlinethat include the filter application are evolved with every passing year. Evolution of newlineActive and passive power filters are the outcome of this necessity. Due to inherent newlinequality and some of the disadvantages of these filters, hybrid structure of filters is used. newlineHybrid Active Power Filter (HAPF) is the one of the most favorable technology to newlinecombat with the harmonic pollution of the voltage and current signals. newlineThe problem with HAPF is that its performance is highly dependent on the newlineparameters of chosen design. Hence, parameter estimation of the HAPF requires accurate newlineestimation of Filter gain, capacitive and inductive reactance components. For estimation newlineof these parameters, many approaches have already been proposed yet the most accurate newlineapproach that provided a very fine tuning of the parameters is still exist in part of newlineresearch. The work presented in this thesis revolve around development of a hybrid newlinemetaheuristic algorithm which is evolved with the fusion of Marine Predator Algorithm newline(MPA) and Sine Cosine Algorithm (SCA). The evolution of this hybrid algorithm is first newlinetested over the set of fixed, Uni and Multimodal benchmark functions. With the analysis newlineon set mathematical functions, efficacy of the hybrid algorithm is established. newline