Uncertainties in measurements and gas path including manufacture tolerance and degradation effects unavoidably influence thrust regulation of gas turbine aero-engines. In this paper, a thrust command scheduling (TCS) controller is proposed based on current measurement precision levels and the improvement of the industrial sensor-based baseline controller, which aims at enhancing the uncertainty tolerance capabilities for a fleet of in-service gas turbine aero-engines. The TCS controller is fulfilled in two steps. A measurement-insensitive thrust mode is selected via random analysis, followed by a two-dimensional thrust command scheduling approach of a family of thrust maps. Industrial baseline controllers with common thrust modes, i.e., low-pressure shaft speed (N1) and engine pressure ratio (EPR) modes are designed as benchmarks. Simulations are conducted on a validated aero-thermal turbofan engine model with publically available uncertainty statistics. Simulation results at the takeoff state on the new and degraded engine fleets reveal that N1 mode is insensitive to measurement uncertainties but owns significant thrust deviation due to degradation effects. Conversely, EPR mode just has the opposite thrust control behavior, compared to N1 mode. The TCS controller regulates the degraded engine fleet with a tight thrust distribution and suppresses the thrust variation of N1 mode via utilizing the remaining N1 margin. Hence, the uncertainty tolerance benefits of the proposed controller are confirmed.

Manufacturing tolerance uncertainties in gas turbine aero-engines are unavoidable, which adversely influence the thrust control performance of newly produced aero-engines. However, classic sample-based uncertainty quantification approaches are usually computationally intensive. In this paper, to consider the uncertainties in the thrust control design phase in advance, a polynomial chaos expansion-based uncertainty model (PCEUM) using a sparse regression method is proposed to get the accurate probability distribution of thrust regulation performance and other concerned engine variables at a decreased computational burden. In PCEUM, interested engine parameters are initially expressed as linear combinations of several orthogonal polynomials, whose weighting coefficients are solved by a sparse-regression-based method, i.e., orthogonal matching pursuit (OMP). Meanwhile, two classic sample-based uncertainty quantification approaches, (i.e., Monte Carlo simulations (MCS), Latin hypercube sampling (LHS)) and least angle regression (LARS) are set as benchmarks. Numerical simulations on a verified large turbofan engine model at the takeoff state on a desktop computer show that PCEUM costs only 47.06 s at 200 samples to obtain converged probability distributions for interested engine parameters whose errors of mean and standard deviation are within 0.01% and 1%, respectively, compared to MCS at 100,000 samples. Meanwhile, compared to the latter three methods, PCEUM saves 94.5%, 81.2%, and 13.1% of the simulation time, accordingly. Hence, both the accuracy and speed of the proposed model are guaranteed for the uncertainty assessment of thrust regulation, which provides a promising solution for both conventional and future aero-propulsion system.