Nonlinear regression by visualization of the sum of residual space appliedto the integrated copolymerization equation with errors in all variables. I. Introduction of the model, simulations and design of experiments

A new model for estimating reactivity ratios using the integrated copolymer ization equation is presented. The model is a general nonlinear least squar es method taking the error in both monomer conversion and monomer fraction into account by a relation between these two variables. Simulations show th at the model is able to predict reactivity ratios successfully. Special att ention is given to experimental design, i.e., at which initial monomer feed ratios the experiments should be performed in order to obtain reliable val ues for the reactivity ratios. (C) 1999 John Wiley & Sons, Inc.