Short-term forecasts gain in accuracy. [Regression technique using ''Box-Jenkins'' analysis]
Box-Jenkins time-series models offer accuracy for short-term forecasts that compare with large-scale macroeconomic forecasts. Utilities need to be able to forecast peak demand in order to plan their generating, transmitting, and distribution systems. This new method differs from conventional models by not assuming specific data patterns, but by fitting available data into a tentative pattern on the basis of auto-correlations. Three types of models (autoregressive, moving average, or mixed autoregressive/moving average) can be used according to which provides the most appropriate combination of autocorrelations and related derivatives. Major steps in choosing a model are identifying potential models, estimating the parameters of the problem, and running a diagnostic check to see if the model fits the parameters. The Box-Jenkins technique is well suited for seasonal patterns, which makes it possible to have as short as hourly forecasts of load demand. With accuracy up to two years, the method will allow electricity price-elasticity forecasting that can be applied to facility planning and rate design. (DCK)
- OSTI ID:
- 7317924
- Journal Information:
- Electr. World; (United States), Vol. 187:2
- Country of Publication:
- United States
- Language:
- English
Similar Records
Automatic load forecasting. Final report
On-line algorithms for forecasting hourly loads of an electric utility
Related Subjects
POLICY AND ECONOMY
20 FOSSIL-FUELED POWER PLANTS
ELECTRIC POWER
FORECASTING
ENERGY MODELS
LOAD MANAGEMENT
MATHEMATICAL MODELS
CHARGES
CLIMATES
ECONOMICS
ENERGY DEMAND
INFORMATION
PERFORMANCE
PLANNING
MANAGEMENT
POWER
296000* - Energy Planning & Policy- Electric Power
292000 - Energy Planning & Policy- Supply
Demand & Forecasting
290200 - Energy Planning & Policy- Economics & Sociology
200106 - Fossil-Fueled Power Plants- Economics