Operating points of a 300 kW solid oxide fuel cell gas turbine (SOFC-GT) power plant simulator are estimated with the use of a multiple model adaptive estimation (MMAE) algorithm. This algorithm aims to improve the flexibility of controlling the system to changing operating conditions. Through a set of empirical transfer functions (TFs) derived at two distinct operating points of a wide operating envelope, the method demonstrates the efficacy of estimating online the probability that the system behaves according to a predetermined dynamic model. By identifying which model the plant is operating under, appropriate control strategies can be switched and implemented. These strategies come into effect upon changes in critical parameters of the SOFC-GT system—most notably, the load bank (LB) disturbance and fuel cell (FC) cathode airflow parameters. The SOFC-GT simulator allows the testing of various FC models under a cyber-physical configuration that incorporates a 120 kW auxiliary power unit and balance-of-plant (Bop) components. These components exist in hardware, whereas the FC model in software. The adaptation technique is beneficial to plants having a wide range of operation, as is the case for SOFC-GT systems. The practical implementation of the adaptive methodology is presented through simulation in the matlab/simulink environment.
Multiple Model Adaptive Estimation of a Hybrid Solid Oxide Fuel Cell Gas Turbine Power Plant Simulator
