Combined modeling for electrical load forecasting with particle swarm optimization

Electrical power forecasting has been always playing a vital part in power system administration and planning. Inaccurate prediction may generate scarce energy resource wastes, electricity shortages, even power grid collapses. Meanwhile, accurate electrical power forecasting can afford reliable guidance for the creation planning of power and the operation of power system, which is also significant for the industry continuable development of electric power. Although thousands scientific papers address electric power forecasting each year, only a few are devoted to finding a general model for electrical power prediction that improves the performance in different cases. This paper proposes a combined forecasting model for electrical power prediction, and the particle swarm optimization is employed to optimize the weight coefficients in the combined forecasting model. The proposed combined model has been compared with the individual models and its results are promising.