基于模式识别的舰船机械电子设备故障自动监测  

Automatic fault monitoring of ship mechanical and electronic equipment based on pattern recognition

作　　者：周丹[1] 熊建华 李柯 ZHOU Dan;XIONG Jianhua;LI Ke(College of Electrical and Mechanical Engineering,Nanchang Institute of Technology,Nanchang 330044,China)

机构地区：[1]南昌理工学院机电工程学院,江西南昌330044

出　　处：《舰船科学技术》2024年第13期82-85,共4页Ship Science and Technology

基　　金：江西省自然科学基金面上项目(20232BAB202003)。

摘　　要：舰船机械电子设备故障数据量较为庞大,且模式复杂多样,为满足其复杂性的要求,提出基于模式识别的舰船机械电子设备故障自动监测方法,采集舰船机械电子设备运行中的温度、压力、振动等数据作为故障监测的原始数据,计算数据间的相似系数和欧氏距离,结合K均值算法实现数据聚类处理。通过小波包算法对聚类后的数据进行特征提取,将其输入到卷积神经网络中,通过对监测模型进行训练,最终实现对舰船机械电子设备故障自动监测。通过实验分析,该方法与相关人员进行监测的故障情况高度一致,在不同故障类型监测的时间均能够保持在5 ms以内,具有较高的监测效率和监测精准度。The data volume of ship mechanical and electronic equipment faults is relatively large,and the patterns are complex and diverse.To meet its complexity requirements,a pattern recognition based automatic monitoring method for ship mechanical and electronic equipment faults is proposed.The temperature,pressure,vibration and other data during the operation of ship mechanical and electronic equipment are collected as the raw data for fault monitoring.The similarity coefficient and Euclidean distance between the data are calculated,and the K-means algorithm is combined to achieve data clustering processing.By using the wavelet packet algorithm to extract features from the clustered data and inputting them into a convolutional neural network,the monitoring model is trained to achieve automatic monitoring of ship mechanical and electronic equipment faults.Through experimental analysis,this method is highly consistent with the fault conditions monitored by relevant personnel,and can maintain monitoring time within 5ms for different types of faults,with high monitoring efficiency and accuracy.

关 键 词：模式识别 舰船机械电子设备 故障监测 K均值算法 小波包算法 卷积神经网络 

分 类 号：TP206[自动化与计算机技术—检测技术与自动化装置]