ALFA: automated load forecasting assistant
Generate an AI Snapshot to get a quick, structured summary of this paper.
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
ALFA, an expert system for forecasting short-term electricity demand is presented and an extensive 10-year historical database of hourly observations of 12 weather variables and a rule base that takes into account daily, weekly, and seasonal variations of load are presented.
Abstract
ALFA, an expert system for forecasting short-term electricity demand is presented. ALFA is in operation at the Energy Management System center at Niagara Mohawk Power Corporation in upstate New York, generating, in real time, hourly load forecasts for up to 48 hours in advance. ALFA uses an extensive 10-year historical database of hourly observations of 12 weather variables and a rule base that takes into account daily, weekly, and seasonal variations of load, as well as holidays, special events, and load growth. A satellite interface for the real-time acquisition of weather data, and the machine-operator interface are also discussed.>