Improvement of Power Quality Using Reduced Switch Count Eleven-Level Inverter in Smart Grid

Abstract

This article proposed a hybrid control topology to cascaded multilevel-inverter (CMLI) using 11-level for the grid-tied photovoltaic generation system. The proposed hybrid technique combines an archerfish hunting optimiser (AHO) and Spike Neural Network (SNN); hence, it is called the AHO-SNN technique. The proposed control technique is to maintain the regulation of power or maximal energy conversion of the solar subsystem and reduce total harmonic distortion (THD). Here, the AHO is utilised to construct the optimum control signal dataset. The best control signals are estimated using a data set performed by SNN. The resultant control signals will regulate insulated-gate-bi-polar-switches (IGBT) of Cascaded MLI. The proposed AHO-SNN control topology determines the converter switching states by constructing the operating modes of the generation system. The system parameter changes and outside disturbances are optimally minimised with this control strategy. The proposed AHO-SNN control is done in MATLAB platform, and it evaluated their performance by using existing methods, like War Strategy Optimization Algorithm (WSO) Fuzzy Wavelet Neural Network (FWNN), Side-Blotched Lizard Algorithm (SBLA), Adaptive Neuro-Fuzzy Inference System (ANFIS), and Memetic Fire-Fly Algorithm (MFA). The result shows that the proposed approach based on THD is less than existing approaches.