基于支持向量机的直升机旋翼不平衡故障分类研究

摘要

A method is presented for diagnosis of helicopter rotor unbalance by using support vector machine (SVM).A diagnosis model based on SVM is established and a model test is conducted on the rotor unbalance composed of blade mass unbalance, blade pitch unbalance and trailing edge flap unbalance.The vibration signals from the rotor test are acquired and analyzed.The SVM-based diagnosis model is employed to identify the rotor unbalance faults and compared with the diagnosis model based on RBF neural network.The results show that the SVM-based diagnosis model is effective in diagnosing the unbalance faults of helicopter rotor with small samples.%提出一种利用支持向量机进行直升机旋翼不平衡故障诊断的方法,建立了用于直升机旋翼不平衡故障识别的支持向量机诊断模型,进行了直升机旋翼不平衡故障模拟试验,分别采集了在旋翼配重不平衡、桨距不平衡、后缘调整不平衡和正常状态下的试验台体振动信号,并对其进行了功率谱分析.采用基于支持向量机的诊断模型对旋翼不平衡故障进行了故障分类识别,并与基于径向基神经网络的诊断模型进行了故障识别效果对比.结果表明基于支持向量机的诊断方法在小样本条件下,对旋翼不平衡故障具有良好的识别能力.