基于PCA的船舶机舱辅机监控系统设计  

Design of monitoring system for marine engine-room auxiliaries based on PCA

作　　者：武茂浦 张国政 卓杰 刘彩云 WU Mao-pu;ZHANG Guo-zheng;ZHUO Jie;LIU Cai-yun(CSIC Information Technology Co.,Ltd,Lianyungang 222006,China)

机构地区：[1]中船重工信息科技有限公司

出　　处：《哈尔滨商业大学学报（自然科学版）》2019年第6期735-739,共5页Journal of Harbin University of Commerce：Natural Sciences Edition

摘　　要：引入主元分析方法,通过对PCA的建模计算和故障检测方法的研究,实现原始数据的降维检测,提升设备状态监测的灵敏度,基于BP神经网络和专家知识进行故障信息的识别诊断及辅助决策.实验表明,该系统能够为机舱辅机设备智能综合管控提供支撑,实现辅机设备状态监测、故障诊断和辅助决策功能,显著提高机舱设备运维效率,快速诊断出发生故障位置以及发生故障原因,从而及时采取有针对性的维修措施,提高船舶机电设备的维修效率,减少船舶设备的故障率.In this paper,principal component analysis(PCA)was introduced to reduce the dimension of original data and improve the sensitivity of equipment condition monitoring.The BP neural network and expert knowledge were used to diagnose fault information.Experiments showed that the system can provide support for intelligent integrated management and control of engine room auxiliaries,and realized the functions of condition monitoring,fault diagnosis and assistant decision-making of auxiliary equipment.It could significantly improve the efficiency of engine room equipment operation and maintenance,quickly diagnose the fault location and causes,so as to take timely targeted maintenance measures,improve the efficiency of ship mechanical and electrical equipment maintenance,reduce the probability of ship failure.

关 键 词：机舱辅机 监控系统 主元分析 降维 BP神经网络 故障诊断 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]