Voltage Profile Based ANN Strategy for Optimal Distributed Generation Placement in Distribution Networks

This paper presents an efficient and scalable methodology for determining the optimal placement of distributed generation (DG) units in radial distribution systems using artificial neural networks (ANN). The proposed strategy aims to enhance voltage profiles and overall system performance while minimizing computational effort. Unlike traditional optimization techniques that require iterative power flow simulations, this approach leverages an ANN trained on voltage profile data from the IEEE 33-bus system to predict optimal DG locations across multiple configurations. After training, the model is validated on a structurally different IEEE 34-bus system to assess generalization capabilities. The ANN achieves 1 0 0% classification accuracy in both training and test cases. Additionally, the method successfully identifies configurations that improve voltage stability, with deviations reduced by up to 4.2% in selected scenarios. The results confirm that the ANN-based method can efficiently classify and prioritize DG placement without exhaustive simulation, offering a promising solution for real-time planning in smart distribution grids.