Identification of heat exchange process of boiler equipment on the basis of perturbed data

The paper considers the process of non-stationary heat conduction and highlights the problem of finding the coefficients of convective heat transfer between the working medium and the inner surface of the boiler drum wall, as well as between the outer heat-insulated surface and the environment. It presents the results of measuring the temperature of the external surface of the boiler unit, which was obtained taking into account the measurement errors. The control of the unobservable heat transfer coefficients aims to prevent the deformation of the drum metal during heating and to provide intensification of the steam and water treatment process. The study employs the structural theory of distributed systems for the purpose of modeling temperature distributions. The issue is thus reduced to the search for the optimal control of an ensemble of trajectories with interval uncertainties, given the information regarding the boundaries of the range of possible temperature distribution. The parameterization of the sought coefficients in the class of constant functions, along with an examination of alternative properties of optimal temperature distributions, gives rise to the problem of parametric optimization formulated with respect to the entirety of temperature realizations across the extended set of arguments. The solution to the problem ensures the identification quality for the most unfavorable trajectory. The results of solving the problem on the basis of noisy experimental data obtained on the boiler equipment of a gas turbine plant demonstrate the efficacy of the method.