ELECTRICITY PRODUCTION FORECASTING FOR BAC LIEU PROVINCE USING DEEP LEARNING NEURAL NETWORKS

Abstract

Electricity system planning is an important requirement in actively building and developing the infrastructure of the power system. It not only supports to meet the increasing requirements of customers, especially for demands of industrial development, but also brings benefits in business, transmission, distribution, and power supply. The electricity indicator forecasting tool plays an important and decisive role for feasible and reliable planning follows the eco-social development. This study aims to propose a solution to apply the long short-term memory (LSTM) neural network to forecast the electricity production and revenue of Bac Lieu province by 2050. This forecasting model is compared with the linear mathematical forecasting model, and the forecasted values of electricity production in the Project of electricity development planning of Bac Lieu province in the period 2016-2025 with a view to 2035. The forecasting results on electricity production of Bac Lieu province is consistent with past trends of the electricity production of developed countries in Europe. And the extension of using this model for other forecasting applications is feasible.