Forecasting Fossil Fuel Energy Consumption for Power Generation Using QHSA-Based LSSVM Model

Energies

Wei Sun¹  Yujun He³  Hong Chang² 

[1] School of Economics and Management, North China Electric Power University, Baoding 071003, Hebei, China;Key Laboratory of Advanced Control and Optimization for Chemical Processes, East China University of Science and Technology, Shanghai 200240, China;Department of Electronic & Communication Engineering, North China Electric Power University, Baoding 071003, Hebei, China; E-Mail:

关键词: fossil fuel energy forecasting;    power generation;    LSSVM;    quantum harmony search algorithm (QHSA);   

DOI  :  10.3390/en8020939

来源: mdpi

【 摘 要 】

Accurate forecasting of fossil fuel energy consumption for power generation is important and fundamental for rational power energy planning in the electricity industry. The least squares support vector machine (LSSVM) is a powerful methodology for solving nonlinear forecasting issues with small samples. The key point is how to determine the appropriate parameters which have great effect on the performance of LSSVM model. In this paper, a novel hybrid quantum harmony search algorithm-based LSSVM (QHSA-LSSVM) energy forecasting model is proposed. The QHSA which combines the quantum computation theory and harmony search algorithm is applied to searching the optimal values of σ and C in LSSVM model to enhance the learning and generalization ability. The case study on annual fossil fuel energy consumption for power generation in China shows that the proposed model outperforms other four comparative models, namely regression, grey model (1, 1) (GM (1, 1)), back propagation (BP) and LSSVM, in terms of prediction accuracy and forecasting risk.

【 授权许可】

CC BY   
© 2015 by the authors; licensee MDPI, Basel, Switzerland.