【作者】 王凯；

【导师】 陈立家；

【作者基本信息】 武汉理工大学 ， 交通信息工程及控制， 2021， 硕士

【摘要】 我国提出建设“交通强国”的目标,以前沿关键科技研发为引领,智慧交通为主攻方向,形成智慧引领的行业发展模式,全面支撑交通发展。为此,交通部提出交通新基建的意见,以数字化、网络化、智能化为主线,促进交通运输提效扩能,加快建设交通强国的总体目标。在交通新基建中,打造智慧航道、智慧港口也是建设的主要内容,其中包括建立航道地理测绘、航行水域气象、水文监测等基础设施以及适应智能船舶的岸基设施等。而在目前的海事管理系统内部,各部分相对独立,关系复杂,缺乏信息资源整合管理,影响了海事业务的集中分析管理。由于基础技术平台的缺乏,水文气象等环境的历史数据也没有投入到船舶交通管理的分析决策中。因此从智慧海事监管系统的设计方向进行研究,提出通过水上交通系统的数字孪生体的构建,实现智慧海事中的监管功能。对于水上交通系统数字孪生体的研究分为水上通航要素感知、船舶交通数学模型和水路环境三维建模三部分进行研究:在水上通航要素感知中,通过对AIS和雷达信息的分析分别还原出船舶轨迹信息,融合AIS和雷达信息关联轨迹;根据CCTV视觉信息识别水上船舶目标,定位船舶动态,结合AIS和雷达信息,确定船舶的动静态参数信息及视觉特征;同时通过传感器设备对风浪等环境信息进行采集,深度挖掘了水上交通感知设备的数据信息。在船舶交通数学建模过程中,通过建立船舶运动模型模拟船舶在真实水上环境作用下的运动响应,结合复杂水域环境下的船舶行为模型实现对水上交通态势的仿真及推演,从而建立水上交通的数学模型。在水路三维建模方面,通过无人机对地形复杂区域采集多视角影像,进行三维重建;对卫星影像图进行图像语义分割,提取建筑轮廓,进行程序化建模,形成水上环境三维模型,为数字孪生模型的可视化提供支持。通过实时通航信息感知、船舶交通数学模型及水路环境的三维建模之间的有机结合,构建水上交通的数字孪生模型,将真实水上交通的动态实时映射到数字空间中。基于该数字孪生模型实现了对水上交通系统的全面感知,预测推演,结合二维、三维可视化技术实现了智慧海事下的海事监管功能升级,包括水上交通态势实时监控、推演、船舶交通事故回放、分析等功能。更多还原

【Abstract】 Our nation puts forward the goal of building national strength in transportation,guided by cutting-edge key technology research and development,take smart transportation as the main direction,forming smart-led industry development pattern to fully support transportation development.To this end,the Ministry of Transportation put forward the opinions on new transportation infrastructure,focusing on digitization,networking,and intelligence to promote the efficiency and capacity expansion of transportation and accelerate the overall goal of building national strength in transportation.In the new transportation infrastructure,building smart waterways and smart ports is also the main content of the construction,including the establishment of waterway geographic mapping,navigational waters meteorology,hydrological monitoring and other infrastructure,and shore-based facilities for smart ships.At present,the internal structure of maritime management system is complicated,and the lack of integrated management of information resources,which reduced the efficiency of the centralized analysis and management of maritime business.Due to the lack of a basic technology platform,historical environmental data such as hydrology and meteorology has not been used in the analysis and decision-making of ship traffic management.This thesis conducts research on the design of smart maritime supervision system,and realizes the establishment of a smart maritime supervision system through digital twins.The research on the digital twin of the water transportation system is divided into three parts: traffic element perception,traffic mathematical model and threedimensional modeling of the waterway environment:In the aspect of traffic status perception,the ship trajectory information is restored and fused through the analysis of AIS and radar information,respectively;based on CCTV visual information to identify ship targets,combine AIS and radar information to determine the ship’s dynamic and static parameter information and visual features;at the same time,environmental information such as wind and waves are collected through sensors.The information of the waterway traffic sensing equipment is deeply excavated.In the process of mathematical modeling of ship traffic,the ship motion model is established to simulate the motion response of the ship under the action of the real water environment,combined with the ship behavior model in the complex water environment to realize the simulation and deduction of the water traffic situation,thereby establishing the mathematics of waterway traffic model.In the procedural of three-dimensional modeling,the drone is used to collect multi-view images on complex terrain areas and perform three-dimensional reconstruction;perform image semantic segmentation on satellite image maps,extract building outlines,and perform programmatic modeling to form a three-dimensional model of the water environment,provide support for visualization of digital twins.Through the organic combination of real-time traffic information perception,water traffic mathematical model and three-dimensional modeling of the water traffic environment,a digital twin model of water traffic is constructed to map the dynamics of real water traffic to the digital space in real time.Based on the digital twin model,a comprehensive perception of the water traffic system,prediction and deduction,combined with two-dimensional and three-dimensional visualization technology to achieve real-time monitoring,deduction,ship traffic accident playback,analysis and other functions of the water traffic situation.更多还原