Multilevel Inverter Real-Time Simulation and Optimization Through Hybrid GA/PSO Algorithm

TL;DR

This work develops a real-time optimization framework for multilevel inverters by integrating a hybrid Genetic Algorithm–Particle Swarm Optimization (GA-PSO) to determine optimal firing angles. The objective combines harmonic minimization and voltage regulation, expressed as $OF = THD(%) + K_v |220 - V_{rms}|$, and the approach is validated on 7- and 11-level inverters in Matlab/Simulink, achieving real-time updates without offline look-up tables. Compared with an ANN-based method, the GA-PSO framework demonstrates superior THD and RMS performance, notably achieving THD as low as $7.47\%$ for the 11-level case, and showing robustness to DC-link and line-resistor variations. The results indicate a scalable, real-time control strategy applicable to multilevel inverters with varying levels, enabling improved power quality in renewable energy systems.

Abstract

This paper presents a new real-time intelligent optimization algorithm to minimize the voltage harmonics of a multilevel inverter. Hybrid Genetic algorithm /Particle swarm optimization algorithm is employed in a real-time simulation to identify the best fire angels of the multilevel inverter to eliminate any destructive effect, such as dc voltage variations and changes in line and dc-link resistors. The dual objective function of harmonic minimization and voltage regulation is considered in this real-time simulation. This approach can be applied to any multilevel inverter with various numbers of levels. The validity of the proposed algorithm is proven by real-time simulation of seven and an eleven-level inverter.

Figures (20)

• Figure 1: Flowchart of The Proposed Algorithm
• Figure 2: Real-Time Objective Function Value
• Figure 3: THD and RMS Voltage Value in Real Time
• Figure 4: Fire Angles of Seven-Level Inverter
• Figure 5: Output Voltage After Regulation (V)