Adaptive neural network approach to one-week ahead load forecasting

T. M. Peng, N. F. Hubele, G. G. Karady

Research output: Contribution to journalArticle

84 Citations (Scopus)

Abstract

A new neural network approach is applied to one-week ahead load forecasting. This approach uses a linear adaptive neuron or adaptive linear combiner called 'Adaline.' An energy spectrum is used to analyze the periodic components in a load sequence. The load sequence mainly consists of three components: base load component, and low and high frequency load components. In load forecasting, the weight rector is designed to make the output sequence, the forecasted load, follow the actual load sequence; it also has a minimized Least Mean Square error. This approach is useful in forecasting unit scheduling commitments. Mean absolute percentage errors of less than 3.4 percent are derived from five months of utility data, thus demonstrating the high degree of accuracy that can be obtained without dependence on weather forecasts.

Original languageEnglish (US)
Pages (from-to)1195-1203
Number of pages9
JournalIEEE Transactions on Power Systems
Volume8
Issue number3
DOIs
StatePublished - Aug 1993

Fingerprint

Neural networks
Mean square error
Neurons
Scheduling

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

Adaptive neural network approach to one-week ahead load forecasting. / Peng, T. M.; Hubele, N. F.; Karady, G. G.

In: IEEE Transactions on Power Systems, Vol. 8, No. 3, 08.1993, p. 1195-1203.

Research output: Contribution to journalArticle

Peng, T. M. ; Hubele, N. F. ; Karady, G. G. / Adaptive neural network approach to one-week ahead load forecasting. In: IEEE Transactions on Power Systems. 1993 ; Vol. 8, No. 3. pp. 1195-1203.
@article{fe876a420b4a4a2084c7c3007a8bc6ea,
title = "Adaptive neural network approach to one-week ahead load forecasting",
abstract = "A new neural network approach is applied to one-week ahead load forecasting. This approach uses a linear adaptive neuron or adaptive linear combiner called 'Adaline.' An energy spectrum is used to analyze the periodic components in a load sequence. The load sequence mainly consists of three components: base load component, and low and high frequency load components. In load forecasting, the weight rector is designed to make the output sequence, the forecasted load, follow the actual load sequence; it also has a minimized Least Mean Square error. This approach is useful in forecasting unit scheduling commitments. Mean absolute percentage errors of less than 3.4 percent are derived from five months of utility data, thus demonstrating the high degree of accuracy that can be obtained without dependence on weather forecasts.",
author = "Peng, {T. M.} and Hubele, {N. F.} and Karady, {G. G.}",
year = "1993",
month = "8",
doi = "10.1109/59.260877",
language = "English (US)",
volume = "8",
pages = "1195--1203",
journal = "IEEE Transactions on Power Systems",
issn = "0885-8950",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "3",

}

TY - JOUR

T1 - Adaptive neural network approach to one-week ahead load forecasting

AU - Peng, T. M.

AU - Hubele, N. F.

AU - Karady, G. G.

PY - 1993/8

Y1 - 1993/8

N2 - A new neural network approach is applied to one-week ahead load forecasting. This approach uses a linear adaptive neuron or adaptive linear combiner called 'Adaline.' An energy spectrum is used to analyze the periodic components in a load sequence. The load sequence mainly consists of three components: base load component, and low and high frequency load components. In load forecasting, the weight rector is designed to make the output sequence, the forecasted load, follow the actual load sequence; it also has a minimized Least Mean Square error. This approach is useful in forecasting unit scheduling commitments. Mean absolute percentage errors of less than 3.4 percent are derived from five months of utility data, thus demonstrating the high degree of accuracy that can be obtained without dependence on weather forecasts.

AB - A new neural network approach is applied to one-week ahead load forecasting. This approach uses a linear adaptive neuron or adaptive linear combiner called 'Adaline.' An energy spectrum is used to analyze the periodic components in a load sequence. The load sequence mainly consists of three components: base load component, and low and high frequency load components. In load forecasting, the weight rector is designed to make the output sequence, the forecasted load, follow the actual load sequence; it also has a minimized Least Mean Square error. This approach is useful in forecasting unit scheduling commitments. Mean absolute percentage errors of less than 3.4 percent are derived from five months of utility data, thus demonstrating the high degree of accuracy that can be obtained without dependence on weather forecasts.

UR - http://www.scopus.com/inward/record.url?scp=0027646290&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0027646290&partnerID=8YFLogxK

U2 - 10.1109/59.260877

DO - 10.1109/59.260877

M3 - Article

VL - 8

SP - 1195

EP - 1203

JO - IEEE Transactions on Power Systems

JF - IEEE Transactions on Power Systems

SN - 0885-8950

IS - 3

ER -