Enhanced UPS inverter control using backstepping and fuzzy neural network for improved power quality

Abstract

The rapid growth of sensitive digital infrastructures and automation systems has intensified the demand for uninterrupted and high-quality power delivery. To address this critical need, this paper proposes a novel hybrid intelligent control strategy for uninterruptible power supply (UPS) inverters that integrates backstepping control, fuzzy neural network (FNN) adaptation, and sliding mode gain compensation. The proposed approach ensures superior voltage regulation and robustness under nonlinear and dynamic load conditions while minimizing dependence on predefined system parameters. The backstepping controller establishes the Lyapunov-based stability framework, the FNN adaptively estimates system uncertainties in real time, and the sliding mode gain enhances resilience against external disturbances. This synergistic control integration enables fast dynamic response, reduced harmonic distortion, and improved system efficiency compared to conventional methods. Simulation and experimental validations demonstrate that the proposed controller achieves total harmonic distortion (THD) below 3%, voltage overshoot under 2%, and enhanced transient recovery, thereby ensuring reliable power quality for critical industrial and commercial applications. The study contributes a real-time feasible, adaptive, and robust UPS inverter control architecture, marking a significant advancement in intelligent power electronics for resilient energy systems.