Enhancing Electricity Consumption Forecasting using Hybrid ANN-ANFIS Models for Smart Grid Applications

Forecasting electricity consumption is a critical task for efficient energy management and for the integration of renewables. The present work uses a combination of two state-of- art techniques, Artificial Neural Networks (ANNs) and Adaptive Neuro-Fuzzy Inference System (ANFIS), to produce more accurate predictions. We develop and compare several forecasting models using four years of data from an institute which employs solar, diesel generator and hydel power. The results reveal that ANFIS does better than ANN with lower Mean Absolute Percentage Error and Mean Squared Percentage Error. The research paper showed that ensemble methods boost electricity consumption prediction outcomes, and they can be by utility companies in their smart grid applications. This research is an effort to make a step forward in data-driven forecasting to meet the growing demand for energy in a sustainable manner and by applying machine learning methods which should maintain data forecasting as needed in planning efforts.