基于部件特性未知的航空发动机故障诊断技术

以航空发动机部件通用特性为基础运用单纯形优化方法,发展了一种预测发动机部件特性的自适应模型方法。并以此为基础运用气路分析原理,选取表征发动机性能参数变化的症兆变量和测量参数,建立了发动机故障诊断小偏差方程。诊断结果表明:症兆变量的选取和相对变化值大小对故障诊断误差有决定性的影响,其值的选取应以门限值1/3为最佳;故障诊断模型症兆变量与实际故障症兆变量变化值相对误差控制在5%内,且不会对故障类型的判断产生影响;故障诊断模型在已知的检验结果内无误诊现象。

Fault Diagnosis for Aeroengines with Components' Characteristics Unavailable

A method was developed to simulate turbofan component characteristics.Nonlinear simplex optimization was used while keeping the deviation of engine performance and working process parameters a minimum as objective function and component coupling coefficients as optimized parameters.A fault diagnosis equation for the twin-spool mixed flow turbofan engine was established based on the simulation and gas path analysis.Investigation shows that the symptom parameters selection and their relatively deviations exert great influence on the diagnostic accuracy.The symptom parameters and their relative deviations should be kept not greater than one third of their threshold values.Relative errors between the deviation of the real fault symptom parameters and of the fault derived from the model are controlled less than 5%.Inaccurate fault isolation does not appear while the developed model was applied to the diagnosis of the available fault cases.