سال انتشار: ۱۳۷۴

محل انتشار: دهمین کنفرانس بین المللی برق

تعداد صفحات: ۱

نویسنده(ها):

Askarian Abyaneh – Amirkabir Univ of Tech
Jahanbakhash – Amirkabir Univ of Tech
Menhaj – Amirkabir Univ of Tech

چکیده:

Hour by hour load forecasting in the frame of medium term can be a suitable parameter for estimating load and energy. Also this forecasting is a good base for short term load forecasting. In this paper a new method for short and medium term load forecasting (hour by hour) using neural network is given. To estimate load in each hour of day, it is necessaty to obey approach of daily load variation and maximum and minimum load of the day under consideration. For classification of various days to estimate load hour by hour in medium term Kohnen neural network being unsupervised one is used. Also for short term load forecasting perseptron neural network is used. At the final section of the paper some results for hour by hour load forecasting of Iran power system are given which shows the method has been satisfactory.