Machine Learning Based Control Techniques for Cascaded H-Bridge Multilevel Inverters to Improve Harmonic Performance

The demand for efficient and intelligent power conversion systems has increased significantly with the rapid growth of renewable energy systems, electric vehicles, and portable power electronics. Conventional inverter systems often suffer from high total harmonic distortion (THD), poor voltage regulation, and limited adaptability to dynamic load conditions. This paper presents an artificial intelligence (AI)-assisted control strategy for a cascaded H-bridge multilevel inverter (CHB‑MLI). Machine learning techniques including convolutional neural networks (CNN), K‑nearest neighbors (KNN), and recurrent neural networks (RNN) are explored for waveform analysis, adaptive pulse width modulation (PWM) tuning, load classification, and predictive fault detection. Simulation and experimental results demonstrate that the proposed AI-assisted method significantly reduces harmonic distortion and improves voltage stability compared with traditional modulation strategies. The study confirms that AI-based control approaches can enhance the performance, efficiency, and reliability of modern inverter systems used in renewable energy, smart grids, and industrial power applications.