SummaryKnowledge from measuring digital simulated processesControling industrial processes demands knowledge of influencing factors. Measurement of those factors is often costly or difficult, sometimes impossible.Constructing a model of such a process may be the answer. One checks the model against the factors that can be measured and uses the results of calculations for factors that cannot be measured.These models should preferably be based on theoretical knowledge of the process, if available. Otherwise statistical models, correlating measurements without explaining them, can be used. When subsystems of the process are sufficiently well known, but interaction and autocorrelation are important, simulation models have proved to be very useful, especially when programmed on digital computers.Some illustrations are given of a heavy plate mill. Slab heating furnaces, mills and shears are treated individually and as components of the whole process.
Read full abstract