Artificial Neural Network–Based Optimization of Solar Photovoltaic and Battery-Integrated Unified Power Quality Conditioner for Microgrid Performance

Abstract

The paper represents a new artificial neural network (ANN)-based control strategy for a unified power quality conditioner (UPQC) with a solar photovoltaic (PV) array and battery energy storage system (BESS) to improve power quality in microgrids. The system utilizes a five-level Cascaded H-Bridge inverter with Sinusoidal Pulse Width Modulation and works under different dynamic conditions such as voltage sag, swell, and harmonic distortion. Simulation results validated that the ANN-based controller performs more as compared to traditional PI controllers by a wide margin, minimizing total harmonic distortion to 2.81% in source current and 2.97% in load voltage. The ANN controller offers intelligent adaptability to nonlinear and unbalanced load conditions, enhancing voltage regulation, current compensation, and harmonic filtering. The coupling of solar PV and BESS improves grid stability and sustainability of energy, while the ANN-based UPQC provides uninterrupted and superior quality power supply.

Cite this article as: K. K. Bharati, S., Y. Shekhar, V. K. Tewari, and O. Singh, “Artificial neural network-based optimization of solar photovoltaic and battery-integrated unified power quality conditioner for microgrid performance,” Electrica, 2026, 26, 0009, doi: 10.5152/electrica.2026.25009.