基于支持向量机的船用低速柴油机换气系统的故障诊断  

Classification of Simulated Faults in Diesel Engine Ventilation System Based on Support Vector Machine

作　　者：曾存 胡以怀[1] 杨雅钧 胡光忠 李方玉 ZENG Cun;HU Yihuai;YANG Yajun;HU Guangzhong;LI Fangyu(Merchant Marine College,Shanghai Maritime Univ.,Shanghai 201306)

机构地区：[1]上海海事大学商船学院,上海201306

出　　处：《内燃机》2020年第5期54-58,共5页Internal Combustion Engines

摘　　要：采用支持向量机对MAN B&W 6S35ME-B9型船用低速柴油机在25%、50%、75%和90%工况下的热工参数进行故障诊断,通过真实热工参数和模拟故障的对比验证,证实了利用支持向量机进行故障诊断的可行性,通过与BP神经网络分类性能比较,说明支持向量机在故障分类中的优越性。The thermal parameters of MAN B&W 6S35ME-B9 marine low-speed diesel engine under 25%,50%,75%and 90%conditions were diagnosed by using support vector machine,and verified by comparison of real thermal parameters and simulated faults.The feasibility of using support vector machine for fault diagnosis is compared with BP neural network classification performance,which shows the superiority of support vector machine in fault classification.

关 键 词：船用柴油机 平均值法 支持向量机 故障诊断 

分 类 号：U664[交通运输工程—船舶及航道工程]