Seven Level Cascade H-Bridge Inverter Modelling using ANN based Selective Harmonic Elimination Technique

Abstract

Because of their efficiency and control methods, inverters have grown in popularity recently as the demand for high-quality power has increased. The main advantage of a multilevel inverter is the ability to generate sinusoidal waveforms. Because of their countless real advantages, these structures have recently gained popularity among industry and utilities. Several in recent years have proposed multilevel inverter topologies, such as a diode-clamped, flying capacitor, and H-bridge multilevel inverters. So, nowadays, cascaded h-bridge inverters are used due to various advantages such as simplicity, fewer components, and so on. With a single-phase cascaded five-level and seven-level inverter, our main objective is to produce an output voltage waveform with excellent quality. A technique known as Selective Harmonic Elimination (SHE) was used to accomplish the stated goal. The inverter is provided with the optimal switching angles during SHE. Artificial Neural Networks (ANN) can carry out this concept. ANN is highly recommended in this scenario because it replaces a memory-intensive look-up table. This method of estimating firing angles for a modulation index is quick and timesaving, increasing the task's effectiveness and efficiency. The ANN model is trained to output the optimal firing angles for a particular modulation index. The Neural Network Fitting Tool in MATLAB is used to train the ANN model. from the simulation results, the seven-level inverter is superior to the five-level inverter, so the hardware implementation of the seven-level cascaded H-bridge is tested for effectiveness and exactness of the results.