Load Forecasting on Special Days & Holidays in Power Distribution Substation Using Neural & Fuzzy Networks

The demand for neural and fuzzy network techniques to predict the increasing load and its application has changed to an ordinary action. However the facts of the real world caused special and exceptional conditions to be created in this network. Like studying the historical background of abnormal behaviors of the network, like holidays or special days, the sudden sale and change in switching operation and changes in customers´habit to decrease the cost of energy [8]-[9] are effective in load prediction. This research investigates the ways of avoiding disharmony and decreasing the errors under the above-mentioned conditions, which will be explained in real examples of current, reactive and active power. For this purpose, we will explain the required conditions for prediction of the next 24 hours, maximum load of the next day and prediction in the next day as fuzzy numbers. Then we will explain the details of project execution, training method and design and then will explain that how the operations will be carried out on special days and holidays and other problems which influence creation of errors in this paper.