基于非线性卡尔曼滤波器的双通道传感器故障检测及隔离方法

Dual-channel Sensor Fault Detection and Isolation Method based on Nonlinear Kalman Filter

航空发动机是典型的非线性系统,传统的基于线性模型的故障诊断算法忽略了高阶非线性特性,对诊断结果有一定的影响。本文建立了发动机改进平衡流形展开模型,提出了基于改进的平衡流形展开模型与容积卡尔曼滤波器相结合的双通道传感器故障诊断及隔离方法,利用双传感器的冗余性实现了快速灵敏地故障检测与隔离。所提方法在CFM-56发动机的模拟机上进行仿真试验,试验结果证明了所提方法的有效性。

The aero-engine is a typical nonlinear system,and the traditional linear model algorithm ignores the higher-order nonlinear characteristics,which has certain influence on the fault diagnosis results.In this regard,an improved equilibrium flow expansion model of the engine is established,and a dual-channel sensor fault diagnosis and isolation method based on the combination of the improved equilibrium flow expansion model and the volumetric Kalman filter is proposed.The proposed method utilizes the redundancy of dual sensors to achieve fast and sensitive fault detection and isolation.At the same time,it can also isolate the engine airline fault and sensor fault,thus avoiding the influence of airline fault on sensor fault diagnosis.The proposed method is simulated on a CFM-56 engine simulator and the experimental results prove the effectiveness of the proposed method.