INTELLIGENT CLASSIFICATION OF RURAL AREA ELECTRICITY CONSUMPTION MANAGEMENT PROFILES USING FUZZY CLUSTERING AND MACHINE LEARNING

Abstract

The rising variability in household electricity usage behaviour needs robust data-driven consumer categorisation systems. This paper suggests using actual monthly electricity consumption data to develop a Fuzzy C-Means (FCM)-based framework for categorising residential electricity consumers. The dataset includes 48 months of power usage data of 1110 residential consumers from Puduppatti Village, Namakkal District, Tamil Nadu, India. The preprocessed and standardised consumption profiles were grouped with FCM for K = 2 to K = 10; clustering quality was evaluated based on the Silhouette Coefficient (SCI) and Davies–Bouldin Index (DBI). The results showed that K = 2 is the most appropriate clustering configuration with SCI of 0.4239 and DBI of 0.9520. The K = 2 solution classified consumers into Low Consumption (LC) and High Consumption (HC) groups. Subsequently, independent test-set evaluation and 10-fold stratified cross-validation were implemented to evaluate five machine-learning classifiers: K-Nearest Neighbour (KNN), Support Vector Machine (SVM), Naive Bayes (NB), Decision Tree (DT), and Logistic Regression (LR). SVM attained the best cross-validation accuracy of 98.56% and independent test-set accuracy of 98.65%, with an AUC of 1.000. The results show that the combination of FCM and supervised learning is successful for categorisation of household power consumption and provides a platform for consumer profiling and tailored energy-management methods.