Mitigating voltage quality issues in Sawla Substation, Ethiopia: A dynamic voltage restorer controlled by neural networks and fuzzy logic

• This study presents an advanced dynamic voltage restorer (DVR) using fuzzy logic controllers (FLCs) and artificial neural networks (ANN) to address power quality issues in the Sawla distribution system. • Simulation results show that the ANN-controlled DVR significantly reduces Total Harmonic Distortion (THD) from 23.92% to 4.94%, meeting IEEE standards, while also effectively mitigating voltage sags and swells. • The findings suggest that the ANN-controlled DVR provides a more stable voltage profile and enhanced system reliability, making it a scalable solution for similar power quality challenges. Ensuring power quality in distribution systems is critical due to voltage sags, swells, and harmonic distortions, which affect system performance and equipment reliability. The Sawla distribution system in Ethiopia experiences high Total Harmonic Distortion (THD), exceeding IEEE Std 519-2014 limits due to nonlinear loads at 15 KV. This study proposes an Artificial Neural Network (ANN) and Fuzzy Logic Controller (FLC)-based Dynamic Voltage Restorer (DVR) to mitigate voltage fluctuations. Using a Pulse Width Modulation (PWM) inverter and d-q transformation, the DVR dynamically compensates voltage variations. ANN-based control outperforms FLC, achieving THD reduction from 23.92% to 4.94%, while the FLC-controlled DVR achieves 12.63% THD, which remains above IEEE limits. MATLAB/Simulink simulations confirm the DVR’s effectiveness in stabilizing voltage and reducing harmonic distortions. This approach enhances power quality and provides a scalable solution for distribution networks.