Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Forecasting the demand for steel

Technical Report ·
OSTI ID:6492082

This study provides forecasts for the weekly demand for cold rolled steel as an input to the production planning and control algorithms being developed. The requirements are weekly forecasts for a horizon of six weeks, and monthly forecasts thereafter for a horizon of six months. One set of forecasts is to be produced for each type of steel, for those types which have the largest demand. Data consisting of the weekly orders for cold rolled steel for various (coded) metallurgical grades over a two-year period was provided by a steel company. The 25 codes with the largest demand over the two-year period were chosen for modeling and ranked in order of greatest demand. Of these, the nine codes with the largest demand had no weekly demands of zero, and one other series had only two zeros. These were modeled using Box-Jenkins techniques. Where Box-Jenkins was inadequate, a new forecasting technique was developed. (PSB)

Research Organization:
Purdue Univ., Lafayette, IN (USA). Purdue Lab. for Applied Industrial Control
OSTI ID:
6492082
Report Number(s):
PUR-102; ON: DE85901828
Country of Publication:
United States
Language:
English