SOLVING THE INVERSE PROBLEM OF CHEMICAL KINETICS FOR A CLOSED NON-ISOTHERMAL REACTOR

A new approach to solving the inverse problem of chemical kinetics based on non-equilibrium experimental data for complex reactions occurring in a closed non-isothermal gradient free reactor of ideal mixing is proposed The described approach is based on a simple property of dynamic models of chemical reactions - the ability to preserve the order of the model when adding any number of linearly-dependent "additional" stages to the original intended reaction mechanism. This artificial extension of the reaction mechanism allows you to keep the number of independent reagents unchanged and reduce the solution of the inverse problem for the original reaction mechanism to the solution of the inverse problem for the extended reaction mechanism. However, the introduction of additional stages can have a significant impact on the relaxation characteristics of the reaction process. To assess this effect, the error introduced by the introduction of additional stages on the dynamics of the reaction was studied Conditions for proximity of the initial and extended models are obtained which impose restrictions on the kinetic parameters of additional stages. It is shown that these conditions can be met if the additional stages are slow enough (they are limiting). To improve the accuracy of solving the inverse problem, the relaxation features (information content) of various time stages of the transition process were also taken into account by allocating sections of fast, medium and slow relaxation. Each of these sections was considered as linear (piecewise linear interpolation), which is. it allows you to calculate reagent concentrations, temperature, and rates of change at any time with sufficient accuracy (not exceeding the measurement error of reagent concentrations and temperature) without using optimization algorithms. As a result of solving the inverse problem, using the described approach, it is possible to estimate the pre-exponents of all speed constants of elementary stages and the intervals of their possible changes for the initial reaction mechanism. Examples of solving the inverse problem for model nonlinear reactions are given. The stability of the method is investigated by applying random noise to experimental data on reagent concentrations and temperature.