Kohonen网络在发动机故障诊断中的应用

文中第一部分研究了利用Kohonen 网络进行发动机故障诊断的特点与算法。并以JT9D发动机为例对算法的有效性进行了检验,23 个故障样本的确诊率达到87% 。文中还以一个典型例子进行了详细分析。文中第二部分提出了利用Kohonen 网络进行故障诊断结果排序的方法,该方法有助于给出简明的诊断结论。文中第三部分提出了利用Kohonen 网络提取故障样本群的代表性样本的方法。该方法对于经验故障方程的建立十分有用。研究结果表明,自组织映射模型程序简单、应用方便,诊断结果不受训练样本的病态影响,特别是能够以非常简单而直观的方式揭示样本群的统计特性,是一种适用于发动机分类故障诊断的较好的、具有特色的神经网络模型

Application of Kohonen Network to Aeroengine Fault Diagnosis

Kohonen Network (Self-organizing Mapping Model) is applied to aeroengine fault diagnosis.The technique of sequencing diagnostic results and extracting typical samples from sample group is presented.Some examples of application of Kohonen Network to JT9D jet engine fault diagnosis are given.It is shown that Kohonen network is simple to compile program and easy to use,can provide correct diagnostic results for the abnormal samples.Especially it can reveal the statistical characteristics of samples clearly and directly.The method of extracting typical samples from sample group is useful for riving the empirical engine fatult equation.