The Integration of Artificial Neural Networks and Particle Swarm Optimization to Forecast World Green Energy Consumption

Abstract

In this study an integrated particle swarm optimization (PSO) and artificial neural network (ANN) is presented for analyzing world fossil fuels, primary energy and green energy consumption. For this purpose, these steps are followed:

STEP 1: In the first step, PSO is applied in order to determine world's oil, natural gas, coal and primary energy demand equations based on socioeconomic indicators. world population, gross domestic product, oil trade movement and natural gas trade movement are used as socioeconomic indicators in this study. Two scenarios are defined for forecasting each socioeconomic indicator in a future time domain:

Scenario I: For each socioeconomic indicator, the polynomial trend line is fitted to the observed data with the highest correlation coefficient value and projected for a future time domain.

Scenario II: for each socioeconomic indicator, a feed-forward back propagation ANN is trained and projected for a future time domain.

STEP 2: In the second step, world green energy consumption is projected based on oil, natural gas, coal, and primary energy consumption using PSO. The best results of the first step are used for future forecasting of world green energy consumption. World green energy consumption is forecasted up to the year 2040.

Keywords:

• artificial neural networks
• forecasting
• fossil fuels
• green energy
• particle swarm optimization

Notes

^aErmis et al., 2007.

^bEIA projection (EIA, 2008).

^aFuture forecasting of each input variable for green energy consumption in future time domain, based on best results of each energy carrier demand estimation in step 1 (Oil_linear, NG_exponential, Coal_linear and PE_linear) according to Scenario I.

^b Future forecasting of each input variable for green energy consumption in future time domain, based on best results of each energy carrier demand estimation in step 1 (Oil_linear, NG_exponential, Coal_linear and PE_linear) according to Scenario II.

^cErmis et al., 2007.

^dEIA projection (EIA, 2008).