Fault-Tolerant Control of CHB Multilevel Inverter Using Adaptive Multi-Objective MPC

Cascaded H-Bridge (CHB) multilevel inverters are widely utilized in renewable energy and high-voltage applications but remain subject to switch faults, causing harmonic distortion, reducing efficiency, and compromising stability. Fault-tolerant methods conventionally relying on hardware redundancy or single-objective optimization cannot scale and adapt to new scenarios. An adaptive multi-objective model predictive control framework (AMO-MPC) is presented in this paper for fault-tolerant operation of a three-phase, seven-level CHB inverter using a dynamically weighted cost function and Recursive Least Squares (RLS)-based parameter estimation. The fault recovery time is 5 ms, Total Harmonic Distortion (THD) is under 15% (IEEE 519 compliant), and efficiency is above 95% while achieving 97% detection accuracy at low signal-to-noise ratios (SNR -20 dB) and ±20% parameter drift. THD is reduced by 23% and recovery time is reduced by 47 % with AMO-MPC compared with sliding-mode and conventional MPC. The study provides software-defined, scalable solutions to improve the reliability, efficiency, and grid integration capabilities of CHB inverters.