基于神经网络的柴油机技术状态主元信息提取方法

Method for Main Information Extraction of Diesel Engine Technical State Based on Artificial Neural Network

在柴油机技术状态监测时,表征其技术状态的特征参数有很多,合理提取状态主元信息是一项关键的任务。分析研究了人工神经网络的信息提取原理和方法。以某型坦克柴油机为例,通过柴油机性能检测试验测取了能够反映柴油机技术状态变化的典型特征,建立了O ja神经网络信息提取模型,提取了柴油机技术状态的主元信息。分析结果表明:提取的主元信息能够反映柴油机技术状态随柴油机使用时间的变化趋势。该方法为坦克柴油机的技术状态监测与故障诊断提供了有效手段。

In monitoring diesel engine,it is a key mission to extract the main information of diesel s technical state from a lot of feature parameters which can reflect the technical state of diesel engine.The theory and methods of artificial neural network in feature extraction are explored in this paper.Taking a certain tank diesel engine for example,the diesel engines on armored vehicles are tested,the typical features that reflect the change of engine technical state are analyzed,a neural network model Oja for fe...