A study on the accuracy of time series modeling of daily power demand

Abstract

The suitability and reliability of the stochastic time series modeling and forecasting approach to load management are examined through actual applications not only to aggregate system load but also to a variety of load types at the 66 22 kV level representing electricity demands in areas of several km2. Daily peak demand forecasts with an accuracy comparable with that of telemetering values have been found to be attainable irrespective of load type, thus establishing the practicality and usefulness of the methodology to operations management and network planning. © 1985.