Authors: Vladimir B. Bokov
Addresses: Postgraduate Department, Army Polytechnic College, ESPE Av. El Progresso, Sangolqui, Ecuador
Abstract: Measurements involve the determination of physical quantities by experiment. In this endeavour, an experimental model will need to specify how measurement system is expected to respond to input data, which is the key to extracting information from the system. The quality of information depends directly on the quality of the model. With this concern novel techniques for model quality improvement have been fashioned. For attaining a high level of comprehensiveness, accuracy and precision, the exact unknown model was approximated simultaneously by available mechanistic and appropriate empirical functions. Adequate modelling was accomplished by employing theoretical and empirical data integration. Herewith, additive and multiplicative approaches were elaborated. The application of developed techniques for sensor model perfection has shown that concurrent multiplicative modelling, in comparison with pure statistical modelling, permits the attainment of less discrepancy in experimental evidence for the whole region of interest for model input variables.
Keywords: experimental design; empirical data; computer experiments; mathematical modelling; information quuality; model quality; data integration; sensor modelling; concurrent modelling.
International Journal of Technology Management, 2007 Vol.37 No.1/2, pp.50 - 71
Published online: 23 Dec 2006 *Full-text access for editors Access for subscribers Purchase this article Comment on this article