为避免船舶设备故障造成的船舶航行风险，设计基于传感器的船舶设备工作状态自动检测系统。利用系统传感器层的多个传感器，采集船舶设备工作时的温度、压力和振动等信号，将采集到的信号传输到应用层，运用流形学习中的t-分布领域嵌入算法降低信号维度。在此基础上采用孤立森林算法，实现船舶设备工作状态自动检测。实验结果表明：该系统选用的传感器具有较好的零点稳定性，且通过降维可以清晰区分正常信号和异常信号，以及异常信号中不同设备故障的信号特征。该系统所得船舶设备工作状态自动检测结果与实际结果的吻合度始终高于96%。
In order to avoid the risk of ship navigation caused by ship equipment failure, an automatic detection system of ship equipment working state based on sensor is designed. Multiple sensors in the sensor layer of the system are used to collect the temperature, pressure, vibration and other signals when the ship equipment is working. The routing node and network coordinator module in the communication layer are used to transmit the collected signals to the application layer. The automatic detection module of the working state of the ship equipment in this layer uses the t-distribution domain embedding algorithm in manifold learning to reduce the signal dimension according to the received signals. On this basis, the isolated forest algorithm is adopted, realize the automatic detection of the working state of ship equipment. The experimental results show that the sensor selected in the system has good zero stability, and the normal signal and abnormal signal and the signal characteristics of different equipment faults in the abnormal signal can be clearly distinguished by dimensionality reduction. The coincidence between the automatic detection results of ship equipment working state obtained by the system and the actual results is always higher than 96%.
2022,44(11): 177-180 收稿日期：2022-01-11
DOI：10.3404/j.issn.1672-7649.2022.11.037
分类号：TP368
基金项目：湖南省科技厅重点研发计划项目（2016WK2023）
作者简介：王梦(1976-)，男，博士，讲师，研究方向为数字电路
参考文献：
[1] 张施令, 何永胜, 宫林, 等. 基于智能图像处理和三维建模技术的特高压换流阀厅用套管运行状态监测方法[J]. 绝缘材料, 2022, 55(1): 87–94
[2] 王鑫, 曾愚, 魏怀灏, 等. 基于改进的Yolo v3模型的设备状态检测方法[J]. 南开大学学报(自然科学版), 2021, 54(6): 7–11
[3] 燕凯, 岳振宇, 高嘉浩. 基于大数据的电力计量设备运行状态评估与故障诊断方法研究[J]. 电子器件, 2019, 42(5): 1095–1098+1121
[4] 余刃, 谢旭阳, 秦法涛, 等. 一种用于旋转设备运行状态无线监测的智能无线振动传感器设计[J]. 核动力工程, 2020, 41(3): 221–226
[5] 周浩, 段发阶, 邵毅, 等. 基于串行通信的传感器自动识别与通用型浮标数据采集系统设计[J]. 海洋科学, 2019, 43(11): 68–75
[6] 曹鹏, 熊圣新, 李建科, 等. 基于5G无线网络的风电机组监控系统组网研究[J]. 船舶工程, 2020, 42(S2): 260–264
