Prediction of Tamil Nadu's Annual Electricity Consumption Using Adaptive Neuro-fuzzy network

In this study, the electricity consumption is forecasted for 10 years from 2021 to 2030 using Adaptive Neuro Fuzzy Inference System (ANFIS) technique from the historical data. State GDP (State Gross Domestic Product), population and income rate are chosen as the independent input variables. In this study, four different model have been developed based on independent input variables for instance type A (State GDP, and population), type B (population and in-come rate), type C (State GDP and income rate) and type D (State GDP, population and income rate). The electricity consumption is the independent output variables for all the models. The different models have been developed to show the impact of independent input variables while predicting the electricity consumption. 3 scenarios are developed according with growth rates of input variables. The simulation results are verified with the National Electricity Plan (NEP) of India.

I. Introduction

Electricity is a significant asserts that influence the economy, political and social security of the nation. Accurate electricity consumption prediction is vital in managing supply side and consumption side of the electricity. The power demand in Tamil Nadu, a state of India has increased around 12% annually from 2010 to 2020 [1]. The electricity utilization has been increased because of the developing population, higher of pay rate, and quick advancement of modern and business areas, more prominent total national output, expanding expectations for everyday comforts and primary changes in the economy of India [2]. Besides the government strategies concerning the improvement of electricity area in world energy markets will assume key parts later on examples of power age and utilization [3].