论文部分内容阅读
针对基于Kalman的故障诊断算法响应速度慢、多故障诊断及非设计点诊断精度低的问题,提出一种基于改进Broyden算法求解方程组的航空发动机气路故障诊断方法。针对涡轴发动机,以模型输出跟踪发动机输出为准则确定3个方程,结合发动机模型中的2个平衡方程,构建气路故障诊断方程组,通过改进Broyden算法求解方程组以获得部件性能退化因子及模型猜值。数字仿真结果表明,所提出的基于Broyden算法求解方程组的航空发动机气路故障诊断方法,在包线内的单故障和多故障诊断稳态误差均小于0.35%,且诊断过程算法单步运行最大耗时小于2ms,具有良好的实时性,远优于Kalman滤波方法,验证了算法的先进性。
Aimed at the low response speed, multi-fault diagnosis and non-design point diagnosis accuracy of fault diagnosis algorithm based on Kalman, a fault diagnosis method of aeroengine gas circuit based on improved Broyden algorithm is proposed. For the turboshaft engine, three equations are determined according to the model output tracking engine output. Combined with the two balance equations in the engine model, the fault diagnosis equations of the gas circuit are constructed, and the system is solved by the improved Broyden algorithm to obtain the component performance degradation factor and Model Guess. The numerical simulation results show that the proposed method based on the Broyden algorithm to solve equations of the aero-engine gas path fault diagnosis method, the single-fault and multi-fault diagnosis steady-state error within the envelope are less than 0.35%, and the diagnostic process algorithm single step maximum Time-consuming less than 2ms, with good real-time, much better than the Kalman filtering method to verify the advanced nature of the algorithm.