A daily peak load forecasting system using a chaotic time series

In this paper, a method for the daily peak load forecasting which uses a chaotic time series and an artificial neural network in a power system is presented. We find the chaotic characteristics of the power load curve and then determine an optimal embedding dimension and delay time. For the load forecast of one day ahead daily peak load, we use the time series load data obtained in the previous year. By using the embedding dimension and delay time, we construct a strange attractor in the pseudo phase plane and the artificial neural network model trained with the attractor mentioned above. The one day ahead forecast errors are about 1.4% for absolute percentage average error.

[1]  Farmer,et al.  Predicting chaotic time series. , 1987, Physical review letters.

[2]  A. Mahmoud,et al.  Load Forecasting Bibliography Phase II , 1981, IEEE Transactions on Power Apparatus and Systems.

[3]  G. Gross,et al.  Short-term load forecasting , 1987, Proceedings of the IEEE.

[4]  Sawada,et al.  Measurement of the Lyapunov spectrum from a chaotic time series. , 1985, Physical review letters.

[5]  Felix F. Wu,et al.  Chaos in a simple power system , 1993 .

[6]  Jianbo Gao,et al.  Direct Dynamical Test for Deterministic Chaos , 1994 .

[7]  Kwang-Ho Kim,et al.  Implementation of hybrid short-term load forecasting system using artificial neural networks and fuzzy expert systems , 1995 .