Electricity Demand Forecasting Using Differential Evolution Algorithm for Tamil Nadu

Abstract

This paper aims to apply the Differential evolution (DE) to forecast the electricity demand for Tamil Nadu based on population, Gross State Domestic Product (GSDP), and Per Capita Income (PCI). The linear and nonlinear models have been applied to forecast electricity demand. The actual data is partially used to attain the best values of the weighting parameters (years from 1980 to 2005) and the remaining data are used for testing the models (years from 2006 to 2017). Three scenarios have been considered (high, average, and low growth) to forecast the electricity demand for Tamil Nadu till the year 2030. Mean Absolute Percentage Error (MAPE) is the estimation key. The results have been validated with the National Electricity Plan (NEP) of India.