A system identification problem of estimating parameters in the profit analysis model for a product

A system identification algorithm with Levenberg-Marquardt method is develo ped to estimate the parameters that were used in the model of profit analys is for a product from the observed expected total profit. The inverse calcu lations are performed based on the simulated exact and inexact observed tot al profit to show the validity of the present study. Moreover, the statisti cal analysis will also be considered here to determine the standard deviati on for the estimated parameters. Results show that the standard deviations of the estimated parameters are decreased as the number of observed profits are increased. Finally the sensitivity analysis is performed to show the v alidity of some results obtained by statistical analysis. (C) 2001 Elsevier Science B.V. All rights reserved.