Design of a Hybrid Meta-Heuristic Optimizer for Modelling a Multi-Level Inverter

Abstract

Meta-heuristic (MH) algorithms have significantly impacted optimization in several technical domains. These algorithms must be implemented in hardware for several technical applications. Hence their performance is essential. Multilayer inverter failure detection is widely applied in High Voltage DC (HVDC) conduction and Industrialized Drives. It uses various meta-heuristic techniques and a NN (Neural Network) as the DM (Decision-Making) mechanism. After the network has been trained for various failure scenarios in the multilevel inverter, the weight and bias parameters are optimized using a MH optimizer to compare the model’s performance. The output of a Multilevel Inverter (ML9LI) supplied by the system is approximated and inferred using a MATLAB-based approach. Features gained from the multi-level inverter, such as positive, negative, and zero sequence voltage and the THD of the output voltage, boost the FD (Fault Detection) ability when using a renewable energy-based power generation system as the basis for the inverter. Particle Swarm Optimization (PSO) and Firefly optimization (FO) are hybridized to form Multi-level Inverter (MLI)-based optimization methods are employed.