Abstract :

High Penetration Of Photovoltaic Distributed Generation (PVDG) Makes Voltage Regulation A Critical Problem In Medium-voltage Feeders. Passing Clouds Change Irradiance By More Than 60 % Within Seconds, Step Voltage Regulators (SVRs) Are Too Slow To Follow, And Smart Inverters Lose Volt-var Capability Because Their Available Reactive Power Changes With The Generated Active Power. This Paper Builds On An Adaptive Voltage Control With Dynamic Voltage Reference, In Which Each PVDG Restores A 10 % Reactive-power Margin And Shifts Its Reference When An Upstream SVR Changes Tap. The Proposed Extension Replaces The Proportionalintegral (PI) Controllers Of The PVDG With Three Artificial Neural Network (ANN) Controllers: ANN-1 Generates The Maximum-power-point Voltage, ANN-2 Replaces The DC-link PI Regulator, And ANN-3 Replaces The Reactivepower PI Regulator. ANN-2 And ANN-3 Are Trained Offline On A Short-horizon Predictive Control Law. Three 300 KW PVDGs, Two SVRs And A Reduced IEEE 34-node Feeder Are Simulated With Fractal Cloud Shading And A Variable Load. The ANN Controller Reduces The Worst DC-link Deviation After Irradiance Steps From 135 V To 33 V And The Settling Time From About 100 Ms To About 30 Ms. The Adaptive Reference Keeps Regulation Capability For 99.8 % Of The Cloudy Period Against 54.6 % With A Fixed Reference. The ANN Regulator Lowers The Voltage Fluctuation Index By 24 % Under Clear Sky And 4 % Under Clouds, Reference Adaptations At Nominal Irradiance Fall From 2520 To 210, And SVR Coordination Cuts Tap Operations From 6 To 2.

Published:

24-10-2026

Issue:

Vol. 26 No. 10 (2026)

Page Nos:

102-110

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite