Author(s): A. W. Jayawardena; D. A. K. Fernando; P. B. G. Dissanayake
Linked Author(s):
Keywords: No Keywords
Abstract: Calibration is an important but difficult task in any mathematical modeling exercise. It requires the determination of the parameters of the model which may be by measurement or estimation. If the mathematical model is physically based, the parameters usually are physically identifiable and therefore measurable at least in theory. However, in situations which involve large numbers of parameters and in problems with process and/or geometrical non-linearities, the determination of parameters by measurement becomes extremely difficult if not impossible. The alternative then is to determine them by estimation which may involve either a trial and error or an optimization approach. The latter is preferred as it is more systematic and objective than the former.
Year: 1997