利用Elman网络良好地逼近非线性能力、容错和泛化能力,根据直升机飞行的实际科目,对待识别的飞行状态预先进行分类,建立Elman神经网络识别模型,验证表明该模型可以较好地识别直升机飞行状态,为直升机传动系统的故障诊断乃至直升机的健康管理提供重要依据。
Elman neural network has the advantages of good close to the nonlinear ability,fault tolerance and generalization ability. According to the dynamic equation of helicopter,an Elman neural network model is established for recognize the flight condition. The computation result shows that the model is competent for recognizing the helicopter flight condition. The model can provide important basis for helicopter transmission system failure forecast and PHM.