Open Access Graduate Research Paper
Prediction of electricity peak demand is an integral part in any electrical utility. Many models were previously proposed for prediction, ranging from short-term, intermediate to long-term prediction. Prediction of electricity peak demand is crucial in avoiding unneeded expenses to upgrade the electrical power supply. It is a continuous process that is still in need of study and improvement. University of Northern Iowa is an educational institution with 10,380 undergraduate students, 66% of whom reside in university housing and 59% of the total electricity demand is purchased from Cedar Falls utilities. The university utilizes a model based on computerized live reading of previous demand that undergoes continuous short-term, intermediate and long-term prediction. The model has succeeded in reducing peak demand. The purpose of the present research is to review the currently used modeling system at UNI, and to test the impact of temperature, humidity, and time variables on peak demand prediction through a Multiple Linear Regression (MLR) short-term model. The MLR model was tested three times, including in each different number of variables, ranging from two to four, with most significant results occurring when four weather and time variables were used.
Year of Submission
Department of Technology
1 PDF file (53 pages)
©2015 Mahmoud Elbokl
Elbokl, Mahmoud, "On the Prediction of the Peak Demand of Electrical Energy Use" (2015). Graduate Research Papers. 3901.