基于LVQ神经网络的柴油发动机故障诊断  

Fault diagnosis of diesel engine based on LVQ neural network

作　　者：王书提 巴寅亮[1,2] 谢鑫[1,2] 

机构地区：[1]新疆农业大学机械交通学院,新疆乌鲁木齐830052 [2]新疆农业工程装备创新设计重点实验室,新疆乌鲁木齐830052

出　　处：《电子设计工程》2015年第17期158-160,163,共4页Electronic Design Engineering

摘　　要：柴油发动机采用电控技术可以优化其动力性、改善燃油使用经济性和控制排放,同时也使得柴油发动机变得更加复杂,增加了故障诊断的难度性。提出了一种基于LVQ神经网络的故障诊断方法,介绍了LVQ神经网络和学习方法。以长城哈佛GW2.8TC柴油发动机电控系统为实验对象,利用金德KT600故障诊断仪采集柴油机的数据流,运用LVQ神经网络建立诊断模型,诊断结果表明利用LVQ神经网络进行故障诊断,效果比较好,具有一定的现实意义。Diesel engine with electronic controlled technology can optimize the power performance, improve the using economy of fuel and curb emissions, but also making diesel engine become more and more complex, increasing the difficulty of fault diagnosis. Proposing a fault diagnosis method based on LVQ Neural Network, introducing LVQ Neural Network and its learning method, taking the electronic controlled system of Great Wall Harvard GW2.8TC diesel engine as the experimental object, collecting the data flow of the engine by kinder KT600 fault diagnosis instrument, using LVQ Neural Network to establish diagnosis model, the diagnosis results show that it is effective to use LVQ neural network to diagnose fault, and has a certain practical significance.

关 键 词：柴油发动机 LVQ神经网络 电控系统 故障诊断 

分 类 号：TN98[电子电信—信息与通信工程]