船用柴油机燃油喷射系统数据可视化分析及预测

Visual analysis and prediction of marine diesel fuel injection system data

燃油喷射系统的故障将直接导致船舶柴油机工作异常,并且船员只有通过驾驶室或集控室的显示器,才能查看数据及报警信息,维修不及时或者错误的情况时有发生。针对这一情形,提出了一种燃油喷射系统故障的可视化分析预测的方法。首先对燃油喷射系统的历史数据进行分析,确认常见的故障类型及相关的属性数据;再将数据整合处理为离散的训练集后,应用基于信息熵的ID3算法,建立决策树故障分类模型。经过试验评估得出分类模型的精度达到94%,满足基本的模型精度要求。最后利用pyecharts技术,进行数据信息和故障类别的可视化展示,便于船舶工作人员随时随地通过移动端查看燃油喷射系统设备运行情况,以保障船舶的航行安全。

The failure of the fuel injection system will directly lead to the abnormal operation of the marine diesel engine,the crew can only view the data and alarm information through the display in the cab or central control room,and the maintenance is not timely or errors occur from time to time.In view of this situation,a method of visual analysis and prediction of fuel injection system failure is proposed in this paper.Firstly,the historical data of the fuel injection system is analyzed to confirm the common fault types and related attribute data.After integrating the data into discrete training sets,the ID3 algorithm based on information entropy is applied to establish a decision tree model of fault classification.The testing results show that the accuracy of the classification model reaches 94%,which meets the basic model accuracy requirements.Finally,the data visualization tool pyecharts technology is used to display the data information and fault categories,and the ship staff are enabled to check the operation status of fuel injection system through mobile terminal anytime and anywhere,so as to ensure the navigation safety.