A Very Short-Term Load Forecasting of Long-Term Fluctuation Components in the Electric Power Demand

It is indispensable to forecast accurately the very short-term load demand to avoid undesirable disturbances in power system operations which deteriorate economical generations. The authors have so far developed a short-term forecasting method by using Local Fuzzy Reconstruction Method, a variant of the methods based on the chaos theory. However, this approach is unable to give accurate forecasting results in case where load demand consecutively exceeds the historical maximum or is lower than the minimum because forecasting is performed by the historical data themselves. Also, in forecasting holidays in summer, forecasting result of weekdays might appear due to similar demand trend. This paper presents novel demand forecasting methods that are able to make accurate forecasts by resolving the above mentioned problems. First, the new method improves the accuracy by extrapolating forecasted transition from the current point. Secondly, to eliminate miss forecast which may be occurred on holidays in summer, historical data are labeled by the information of the day of the week to distinguish similarly behaved weekdays’ load patterns. The proposed methods are applied to 10, 30, and 60 minutes ahead demand forecasting, and the accuracy is improved 10% to 20% compared with the method previously proposed.

[1]  Tetsuro Matsui,et al.  Development of peak load forecasting system using neural networks and fuzzy theory , 1996 .

[2]  Tatsuya Iizaka,et al.  Development of Electric Load Forecasting System using Neural Networks , 2000 .

[3]  Tadashi Iokibe.  Chaos and Prediction , 1995 .

[4]  K. Yukita,et al.  Study of daily peak load forecasting by structured representation on genetic algorithms for function fitting , 2004, 39th International Universities Power Engineering Conference, 2004. UPEC 2004..

[5]  Jiménez,et al.  Forecasting on chaotic time series: A local optimal linear-reconstruction method. , 1992, Physical review. A, Atomic, molecular, and optical physics.

[6]  H. Mori,et al.  Deterministic annealing clustering for ANN- based short-term load forecasting , 2001, PICA 2001. Innovative Computing for Power - Electric Energy Meets the Market. 22nd IEEE Power Engineering Society. International Conference on Power Industry Computer Applications (Cat. No.01CH37195).

[7]  Michael T. Manry,et al.  Comparison of very short-term load forecasting techniques , 1996 .

[8]  W. Charytoniuk,et al.  Very short-term load forecasting using artificial neural networks , 2000 .

[9]  Yasunari Fujimoto,et al.  Local Fuzzy Reconstruction Method for Short-term Prediction on Chaotic Timeseries , 1995 .

[10]  Taro Kawase,et al.  Forecast of Daily Maximum Electric Load by Neural Networks using the Standard Electric Load , 1997 .

[11]  Kazuyuki Aihara,et al.  Forecasting Daily Peak Load by a Deterministic Prediction Method with the Gram-Schmidt Orthonormalization , 1995 .

[12]  Alistair Mees.  Dynamical Systems and Tesselations: Detecting Determinism in Data , 1991 .

[13]  F. Takens.  Detecting strange attractors in turbulence , 1981 .

[14]  Eiichi Tanaka,et al.  Trends of R&D on Short-Term Load Forecasting , 1994 .

[15]  Hiroshi Sasaki,et al.  Development of very‐short‐term load forecasting based on chaos theory , 2004 .