Developing Annual Load Duration Curve Using an Intelligent Technique


AbstractThe paper discusses a method to get the hourly load data from the peak daily load (i.e. to store one value and get 24 values). The method uses one of the intelligent techniques (Artificial Neural Network ANN). The annual load duration (ALDC) is used in different studies, such as power system planning, reliability study . . . etc. In this work the ALDC is used as an example for the application of the proposed method. When the hourly load data for a year are available, it is easy to find the ALDC. In the studies where the ALDC is needed, the load is usually forecasted (future load) where the hourly data are not available. A proposed ANN is explained to overcome this difficulty. The method develops the daily load (24 hours) from the peak load. The required days are encountered, which means that the ALDC is obtained for the required days. Also if there is a missing period, the proposed method can develop that missing period in the data. The data of the Iraqi North Region National Grid (INRNG) for the year 2001 is used to verify the validity of the proposed method. The results of a conventional method are also given.Key words: ALDC, Daily load, Hourly load, ANN.