基于过渡工作过程的航空发动机气路分析方法

Aero-engine Gas Path Analysis Method during Transient Process

目前,绝大多数有关航空发动机气路分析方面的研究均基于发动机的稳态过程,但稳态数据通常难以获得,且有些故障特征仅在过渡工作状态下才能表现出来,因此使用相对更容易获得的过渡态数据能够更全面地对发动机的健康状态进行评估。在气路传感器数目较少的情况下,采用序列工作点方法对大量健康参数进行分析,在增加可用信息量的同时,降低由多工作点方法的平均效应引入的参数估计系统误差;提出间接递归牛顿-拉夫逊法强化非支配分类差分进化算法,以解决发动机大偏差性能退化健康参数估计中的计算收敛性问题。对某型双轴分排涡扇发动机进行气路分析,结果表明:本文方法能够在气路传感器数目有限的条件下,利用发动机过渡态数据实现对大偏差范围内大量健康参数的高效、准确估计。

At present, most of the research on aero-engine gas path analysis is based on the steady-state process of the engine, but the steady-state data is usually difficult to obtain, and some fault features can only be displayed in the transient working state, so the transition state data which is relatively easy to obtain can be used to evaluate the health state of the engine more comprehensively. In the case of a small number of gas path sensors, a large number of health parameters are analyzed by using the sequential operating points method, which can increase the available information and reduce the systematic error of parameter estimation introduced by the average effect of the multiple working points method. An indirect recursive Newton Raphson method is proposed to enhance the non-dominated sorting differential evolution algorithm to solve the problem of computational convergence in the estimation of large deviation performance degradation health parameters of engines. The results show that this method can effectively and accurately estimate a large number of health parameters in a large deviation range by using engine transition state data under the condition of limited number of gas path sensors.