智慧港口海运危险货物集装箱作业安全管理研究  

Research on Safety Management of Marine Hazardous Cargo Container Operation in Smart Port

作　　者：汪侃 汪皓 刘晓雷 许涛 胡倩倩 WANG Kan;WANG Hao;LIU Xiaolei;XU Tao;HU Qianqian(College of Ocean Science and Engineering,Shanghai Maritime University,Shanghai 201306,China;Shenzhen Technology Institute of Urban Public Safety,Shenzhen Guangdong 518024,China)

机构地区：[1]上海海事大学海洋科学与工程学院,上海201306 [2]深圳市城市公共安全技术研究院,广东深圳518024

出　　处：《安全》2024年第9期77-85,共9页Safety & Security

基　　金：国家自然科学基金项目(52001196);上海市“科技创新行动计划”软科学研究项目(20692193100)。

摘　　要：为增强智慧港口建设中海运危险货物作业全流程的安全管理精细化水平,本文基于数字孪生理念针对海运危险货物集装箱作业全流程提出数字孪生五维框架,从物理实体、虚拟模型、孪生数据、信息传输和应用服务等方面开发海运危险货物集装箱数智应急平台。通过构建船载集装箱、岸桥卸箱、场桥抓箱、AGV运箱、外集卡载箱等环节的全方位数字孪生场景,由集装箱体内部安装的传感器提供多源数据,实现危险货物集装箱内状态与数字孪生模型之间的信息交互;在海运危险货物集装箱数智应急平台中内嵌CNN+LSTM深度学习模型,对危险货物集装箱内温湿度等风险参量进行变化趋势预测。结果表明:温度预测值与真实值之间的MSE和RMSE值为19.27和4.39,MAE和MAPE值为7.29%和2.58%,决定系数达到0.9855,表明数字孪生模型对集装箱内部温度变化能够准确预测。这对危险货物在智慧港口作业中的事故预先研判和异常集装箱及时应急处置具有重要作用。In order to enhance the safety management refinement level of the whole process of marine hazardous cargo container operation process in the construction of smart port,based on the concept of digital twin(DT),a five-dimensional DT framework for the whole process of marine hazardous cargo container operation was proposed in this paper.Subsequently,a digital emergency platform for marine hazardous cargo container was developed,which contained the aspects of physical entity,virtual model,twin data,information transmission and application services.By constructing a comprehensive DT-based scenario covering ship-borne container,quay crane unloading container,yard crane capturing container,AGV transporting container,and truck loading container,etc.,the multi-source data is provided by sensors installed inside the container body to realize the information exchange between the status of hazardous cargo container and virtual DT model.The CNN+LSTM deep learning model was embedded in the marine dangerous goods container number intelligence emergency platform to predict the change trend of risk parameters such as temperature and humidity in hazardous cargo container.The results show that the mean squared error and root mean squared error values between the predicted and true temperature values are 19.27 and 4.39,respectively,the mean absolute error and mean absolute percentage error values are 7.29%and 2.58%,respectively,and the coefficient of determination reaches 0.9855,which indicates that the DT model can accurately predict the temperature change inside the hazardous cargo container.The proposed DT system plays an important role in the accident prediction and emergency treatment of hazardous cargo in the smart port operation and the timely emergency handling of abnormal containers.

关 键 词：安全管理 危险货物集装箱 作业管理 智慧港口 数字孪生 

分 类 号：X951[环境科学与工程—安全科学]