Estimating the parameters of chemical kinetics equations from the partial information about their solution

Abstract

The inverse problem of identifying the parameters of sets of ordinary differential equations using experimental measurements of three functions that correspond to some components in the vector solution of a set is considered. A private case that is important for applications of chemical and biochemical kinetics when reduced equations linearly depend on the combinations of initial unknown parameters has been studied. An analysis and the numerical results are presented for two types of sets of chemical kinetics equations, such as the Lotka–Volterra model that describes the coexistence of a predator and a prey and the chemical kinetics equations that model enzyme catalysts reactions, including the Michaelis–Menten equations. The search for unknown parameters is confined to the problem of minimizing a quadratic function. In this case, the reduced differential equations of systems are used instead of their vector solutions, which are unknown in most cases. The cases of both stable and unstable search for unknown parameters are analyzed.