Power Quality Enhancement in Smart Grids Using an Energy Storage System With Brown Bear Optimization and a Boundary‐Integrated Neural Network

ABSTRACT Smart grid (SG) environments are increasingly affected by power quality (PQ) disturbances arising from high penetration of renewable energy sources (RES), nonlinear loads, and fluctuating operating conditions. These difficulties require smart, dynamic control systems that can stabilize voltage and frequency and reduce harmonic distortion. One of the major motivations for this work is the difficulty of coordinating Energy Storage Systems (ESS) within SGs to ensure seamless PQ management under dynamic, uncertain energy profiles. To overcome this obstacle, this paper presents a combined SG scheme comprising photovoltaic (PV), fuel cell (FC), and battery storage systems interconnected via DC‐DC converters and connected to the utility grid. In this architecture, a hybrid Brown Bear Optimization‐Boundary Optimization Integrated Neural Network (BBO‐BINN) is proposed to improve PQ performance. BBO optimizes the control behavior of the ESS to minimize Total Harmonic Distortion (THD) and regulate voltage‐frequency characteristics, whereas BINN predicts PQ disturbances using boundary‐aware learning to support adaptive and proactive grid responses. The approach presents a single optimization‐prediction process that enhances energy management efficiency and reliability of operation. The simulation results show that the suggested framework achieves a high PQ (2.70% minimized THD) and that the BBO‐BINN algorithm has an efficiency of 98.07%, indicating that it is effective in stabilizing and improving SG performance.