Recursive least squares for online dynamic identification on gas turbine engines

Online identification for a gas turbine engine is vital for health monitoring and control decisions because the engine electronic control system uses the identified model to analyze the performance for optimization of fuel consumption, a response to the pilot command, as well as engine life protection. Since a gas turbine engine is a complex system and operating at variant working conditions, it behaves nonlinearly through different power transition levels and at different operating points. An adaptive approach is required to capture the dynamics of its performance.