Evaluating Power Consumption Model and Load Deficit at Different Temperatures Using Clustering Techniques and Presenting a Strategy for Changing Production Management

Abstract

The present study evaluated the consumption model among the subscribers of the power industry at a certain temperature range by focusing on the amount of load deficit and using clustering techniques in data mining. This study an initial dataset including the load demand, power plant reserve, power import, and power export at the range of -12 to -50 degrees centigrade. Then, the optimal algorithm was selected by evaluating three clustering algorithms of Single Linkage, Complete Linkage, and Average Linkage and comparing to K-means algorithm using the key indices of clustering performance. Finally, a power export model was proposed to reduce load deficits and an appropriate solution to compensate for this deficit. The results indicated that the supply and distribution network faced load deficit and caused electric breakdowns due to the increased demand for power consumption. This study identified the desired temperature range and proposed some suggestions for reducing the volume of load such as the reduced amount of export, increase of import, and the increase in power plant production in accordance with capacity.