基于RBF神经网络的航空发动机故障诊断研究  

Fault Diagnosis of Aeroengines Based on RBF Network

作　　者：郭风[1] 王思远[2] 崔红军[1] 

机构地区：[1]中国人民解放军94333部队 [2]中国人民解放军94590部队

出　　处：《航空计算技术》2007年第2期23-25,29,共4页Aeronautical Computing Technique

基　　金：国防预研基金资助项目(98J19.3.2.JB3201)

摘　　要：采用改进算法优化了径向基函数(RBF)网络。针对航空发动机工作条件和结构的复杂性,提出了用RBF网络进行故障诊断的方法,构建了基于RBF网络的多参数航空发动机故障智能诊断模型,并对典型发动机故障进行了诊断。结果表明,RBF网络具有优秀的故障学习能力,采用它进行航空发动机故障诊断是行之有效的,具有较好的应用前景。A RBF network based on improved algorithms is described. An aeroengine fault diagnosis technique based on RBF network is presented for the complexity of operating conditions and structures of aeroengines. And then, an aeroengine intelligent fault diagnosis model is proposed. The classic aeroengine faults are test with the diagnosis model. The practical application results show that the method proposed is accurate and effective, and it is in good agreement with the actual inspections.

关 键 词：RBF网络 航空发动机 故障诊断 智能诊断 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]