Artificial Intelligence -Based Control Strategies for Cascaded H-Bridge Multilevel Inverters: A Review

The increasing demand for compact and reliable portable power supplies has created the need for efficient inverter systems capable of delivering high-quality AC output from DC energy sources. Conventional low-cost inverters often suffer from high total harmonic distortion (THD) and poor efficiency due to their simple design and lack of adaptive control. This paper presents an intelligent control strategy for a cascaded H-bridge multilevel inverter (CHB-MLI) integrated with artificial intelligence (AI)-based optimization techniques. Convolutional Neural Networks (CNNs) are utilized for real-time waveform analysis and THD reduction through adaptive pulse-width modulation (PWM) control. The K-Nearest Neighbors (KNN) algorithm is applied for load classification and optimal switching pattern selection, while Recurrent Neural Networks (RNNs) are employed for predictive load management and fault detection. A hardware prototype was developed using MOSFET-based H-bridge modules, a DSP controller, and a Raspberry Pi 4 as the AI processing unit. Experimental results demonstrated that the proposed AI-assisted CHB-MLI achieved a significant THD reduction from 12.6% to 3.4%, enhanced voltage stability, and improved dynamic response under varying load conditions. The findings validate the effectiveness of integrating AI techniques in inverter control, offering a promising approach for smart and efficient portable power systems. Furthermore, the proposed system has wide-ranging applications in renewable energy systems (RES), electric vehicles (EVs), smart grids, industrial automation, and medical equipment, where high efficiency, reliability, and power quality are essential.