文献类型：期刊文献

中文题名：船舶机电设备故障诊断方法研究

英文题名：Ship's Electrical and Mechanical Equipment Fault Diagnosis Methods

作者：李文星[1]

第一作者：李文星

机构：[1]新乡学院机电工程学院

第一机构：新乡学院机电工程学院

年份：2013

卷号：21

期号：8

起止页码：2093-2095

中文期刊名：计算机测量与控制

外文期刊名：Computer Measurement & Control

收录：CSTPCD;;北大核心:【北大核心2011】；

语种：中文

中文关键词：船舶机电设备;故障检测;数据分类

外文关键词：ship electrical equipment; fault detection; data classification

摘要：当前几乎所有的船舶设备故障检测方法都没有考虑各船舶机电装置在运转状态、管理水平等方面的差异,没有统一的检测模式,不能准确检测所有的突发性故障,故障检测准确率也不高,为了解决这一问题,提出了一种新的船舶机电设备故障诊断方法;运用各种信号传感器等硬件装置采集船舶故障信号数据,形成差异化检测数据环境,方法中引入一种基于最小支持向量机的故障数据分类机制,对船舶机电设备产生的所有故障数据进行有效的分类,形成统一检测环境,去除强冗余,强干扰数据,保证故障数据特征的准确提取,设定合理的阀值进行检测;实验结果表明,该方法能够在干扰因素较多的冗余情况下对船舶机电设备故障进行准确检测,工作性能较好。
Current almost all Marine equipment fault detection method without considering the ship electromechanical device in operation status, management level differences, no unified examination mode, cannot accurately detect all of the sudden fault, fault detection accuracy is not high also. In order to solve this problem, this paper proposes a new ship electrical and mechanical equipment fault diagnosis methods. Use all kinds of signal sensor device such as a hardware fault signal data acquisition of ship, differentiate detection data environment, intro- duce a method based on minimum failure data support vector machine （SVM） classification mechanism, of the ship electrical and mechanical equipment to produce all the failure data for effective classification, form a unified testing environment, remove redundancy, noise data, guarantee the accuracy of the fault characteristics of the data extraction, set a reasonable threshold for testing. The experimental results show that the method can more interference factors in the redundant cases to accurate shipping mechanical and electrical equipment failure de- tection, has good working performance.