航空发动机转速基准建模方法研究

Addressing the current issue of low confidence in fuel flow measurements， which significantly impacts the accuracy of real-time dynamic models in aero-engines， this paper introduces a novel modeling methodology that replaces the traditional fuel reference with a rotational speed reference. This approach uses a rotational speed reference instead of the traditional fuel-based reference. Building upon conventional component-level real-time dynamic model， this method incorporates the measured state parameter of rotational speed as the reference target for the model. Fuel flow is then utilized as an adjusting parameter to align the model rotational speed with the measured rotational speed， indirectly achieving automatic correction of input fuel and thereby enhancing model accuracy. Furthermore， an enhanced multi-constraint Proportional-Integral （PI） control algorithm is introduced to achieve precise rotational speed alignment control. The proposed modeling approach is validated through the modeling of actual ground test data and simulated fuel measurement error cases. The results indicate that the proposed rotational speed reference model achieves significantly lower modeling errors for four key parameters： fan speed， compressor speed， total pressure after the compressor， and total temperature after the turbine， when compared to traditional method. Specifically， the mean absolute errors for these parameters were reduced by 90.0%， 55.6%， 27.1%， 29.0%， respectively. These findings validate the effectiveness of the proposed modeling approach.