Load Transfer Coupling Regression Curve Fitting for Distribution Load Forecasting
- 1 May 1984
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Apparatus and Systems
- Vol. PAS-103 (5) , 1070-1076
- https://doi.org/10.1109/tpas.1984.318713
Abstract
The planning of distribution facilities requires forecasts of future substation and feeder oads. Extrapolation based on a curve fit to past annual peak loads is currently the most popular manner of accomplishing this forecast. Curve fitting suffers badly from data shifts caused by switching as loads are routinely moved from one substation to another during the course of utility operations. This switching contaminates the data, reducing forecast accuracy. A new regression application reduces error due to these transfers by over an order of magnitude. A key to the usefulness of this method is that the amount of the transfer, and its direction (whether it was to or from a substation), is not a required input. The new technique, aspects of computer implementation of it, and a series of tests showing its advantage over normal multiole regression methods are given.Keywords
This publication has 3 references indexed in Scilit:
- Spatial electric load forecasting: A tutorial reviewProceedings of the IEEE, 1983
- A Hierarchical Recursive Method for Substantially Improving Trending of Small Area Load ForecastsIEEE Transactions on Power Apparatus and Systems, 1982
- Some unique signal processing applications in power system planningIEEE Transactions on Acoustics, Speech, and Signal Processing, 1979