用机体振动诊断旋翼失衡故障方法研究

Novel Helicopter Rotor Imbalance Fault Diagnosis Using only Information from Fuselage Vibration

直升机失衡故障通常通过旋翼桨尖运动轨迹和机体 1/转次振动测量进行诊断。为简化诊断 ,本文提出了仅通过机体振动测量 ,不需要桨尖轨迹及其测量设备进行诊断的新方法。证明了旋翼失衡故障空间与机体 N=r× q维振动空间存在一对一映射关系 ,其中 r为用于诊断的振动信号频谱前 r阶谱分量数 ,q为安装在机体上作振动测量的单轴传感器数。通过两个 BP神经网络构造映射关系 ,一个用于故障分类 ,另一个用于故障程度诊断。通过实测机体振动信号频谱前 r阶幅值分量对网络进行训练和测试。失衡桨叶的方位由实测振动频谱一阶分量的相角确定。在三片桨叶的模型旋翼上的实验结果表明 ,所提出的方法可行。

The diagnosis of helicopter rotor imbalance faults, including mass and aerodynamic imbalance traditionally depends on the measurements of the flap and lead /lag track of the rotor blades and the fuselage vibration at 1/rev frequency. In order to simplify the balancing procedure, a novel method is developed and presented in this paper, which makes the diagnosis possible using only information extracted from rotor induced fuselage vibration without any rotor blades track measurements and associated optical devices. In this work, it is proved that there exits an injection of rotor imbalance fault space into rotor induced fuselage vibration space with N=r×q dimensions, where r is the number of the first harmonics of the vibration, and q the number of single axis accelerometers mounted on the fuselage for vibration measurements. The injection is identified by using two BP neural networks, one is for fault classification and the other for the fault magnitude diagnosis. The networks are trained and tested by measured fuselage vibration signal in frequency domain. The location of the unbalanced blade is identified by the initial phase of the fundamental harmonic component of the measured data. The test results of a model rotor with three blades indicate that the method is feasible and effective.