Prediction of Electricity Demand Considering Interactions among Economic Factors Using a Bayesian Least Squares Support Vector Machine Method

With the rapid development of national economy in Fujian Province, the balance of power supply and demand has become an important aspect of power market supporting social and economic development. In this study, the Bayesian least squares support vector machine method is used to forecast electricity demand by Fujian (EDF). The major findings are: (i) The economic data (GDP), primary industry (PI), secondary industry (SI), tertiary industry (TI) and residents’ consumption level (RCL) of Fujian Province has been growing over the years, and it is predicted that it will continue to grow in the next 15 years. (ii) By 2035, the power demand of Fujian Province would reach 2677.55 billion kWh, an increase of 10.3% compared with 2020; (iii) Electricity consumption would increase with economic growth. The application of power demand forecasting model in Fujian Province can provide basis for power enterprises to do well in power supply and power grid planning and demand side management in the future, and better support urban socio-economic development and operation and management of power enterprises.

I. Introduction

Electric power is the material basis for social and economic development, people’s daily life and all walks of life. The development of electric power has provided strong material support for China’s sustained and rapid economic growth; On the contrary, the sustainable development of economy can promote the development of power industry [1], [2]. The shortage of electric power will seriously restrict the sustainable and healthy development of economy. Because electricity is different from other energy sources, it cannot be stored and has no inventory, which can truly reflect the operation status of social economy. Therefore, the demand and supply of electricity for economic development often become the wind vane of social and economic development [3], it is important to study the impact of economic development on electric demand prediction.