Inclusive modelling and parameter estimation of a steam power plant using an LMI-based unknown input reconstruction algorithm

Abstract

A comprehensive dynamic model of a steam power plant, including turbine, governor, and boiler models, is proposed in this paper. The whole turbine-boiler control strategies inter alia turbine following, boiler following, coordinated control, and sliding pressure control modes are also taken into account in the modelling procedure, making the integrated model adaptable for all the steam units. A novel multi-stage algorithm is asserted in the other part of the paper to estimate the obtained model parameters using input-output data from operational tests. Regarding the fact that the mechanical power as the turbine output and one of the synchronous generator (SG) inputs cannot be measured directly, an efficacious unknown input reconstruction (UIR) method is employed in this paper to determine the mechanical power from SG measured signals. The UIR observer in this method is obtained by solving a linear matrix inequality (LMI). Other parameters are identified by applying a Genetic Algorithm (GA) to the measurement data. Finally, the acquired model and parameters are validated by employing the recorded data from several tests, and the results are discussed.