Minimization of THD in Nine Level Cascaded H-Bridge Inverter Using Artificial Neural Network

TL;DR

This paper presents a neural network-based control method to minimize Total Harmonic Distortion in a nine-level Cascaded H-Bridge inverter, demonstrating significant THD reduction through simulation.

Contribution

It introduces a neural network approach for predicting switching angles to effectively reduce harmonics in multilevel inverters, outperforming traditional open-loop control methods.

Findings

THD reduced to about 3% with neural network control

Neural network prediction improves harmonic performance

Simulation results validate the effectiveness of the proposed method

Abstract

Multilevel inverter converts different level DC voltage to AC voltage. It has wide interest in power industry especially in high power applications. In power electronic equipment the major drawback is the harmonics. Several control strategies are available to reduce the harmonic content and the most widely used measure of Total Harmonic Distortion (THD). In this project, the comparison has been made for the open loop and closed loop PI controller and neural network that predict the switching angle in order to reduce the harmonics. The mapping between Modulation Index and Switching angles are plotted for the forward neural network. After the prediction of switching angles the neural network topologies are executed for better result. This technique is applied for any type of multilevel inverter, Cascaded H-Bridge multilevel inverter is chosen. A nine level Cascaded H-Bridge multilevel inverter power circuit is simulated in MATLAB 8.3 simulink with sinusoidal PWM technique. The comparison results reveal that the THD is reduced to about 3% with neural network control compared to open loop control. The results are presented and analyzed.