An Intelligent Ensemble Classifier for the Transformer-User Relationship Identification in Distritbution Networks

This paper introduces an intelligent ensemble classifier for determining the relationship between transformers and users in distribution networks. The classifier combines three classification models with different structures and characteristic: Neural Networks (NN), Support Vector Machines (SVM), and Random Forests (RF). A soft voting mechanism is used to combine the output probabilities from all these three models, selecting the category with the highest probability as the final prediction. By utilizing the predictive capabilities of each model, this approach allows for more robust decision-making, particularly when category differences are minimal. The performance of the proposed ensemble model is validated through user identification tests conducted across three neighboring distribution areas.