Elman回归神经网络在航空发动机故障诊断中的应用

提出了一种基于Elman神经网络的航空发动机故障智能诊断方法。以在某型发动机地面定检状态下实测的数据作为样本数据,建立了航空发动机故障诊断模型,利用该模型成功地对实测发动机参数进行了诊断。研究结果表明:该方法对权值初始值选取、隐含层节点数选取和输入样本规范化处理等不敏感,具有学习速度快、诊断精度高等优点。

Application of Elman Neural Network in Aeroengine Fault Diagnosis

A method of aeroengine intellengence fault diagnosis is developed based on the Elman neural network.The fault diagnosis model of aero engine is set up by using the actual measurement of an engine at the ground periodic check condition as sample data.The model is successfully applied to diagnosis some actual engine parameter.The results show that the method is not sensitive to the selection of initial weight and neuron numbers of hidden layer,as well as standard disposing of sample data,and it has the advantages of faster learning and higher diagnosis precision.