面向通航环境数字化灰信息下的北极航道船舶冰困风险预警研究

Since the thickness and coverage area of Arctic ice decrease year by year along with the Global warming, the navigation of Arctic Sea routes is expected to become a reality. However, the risk warning of ship besetting in ice is the key point and difficult problem to achieve the safety of Arctic navigation. Firstly, this project studies the grey information characteristics and information fusion technology of the elements of the Arctic Sea routes, facing the digitalization of information resources of navigation environment. Then, on the view of three dimension features, including time, space location, and spatial and temporal characteristics, the group of digital grey information correlation models is established considering the delay and periodicity of the risk factors, to research the mapping relation between risk factors of Arctic Sea routes environment and navigation states of ship, as well as the recognition of risk factors and the division of navigation states. Thirdly, multi variable grey rolling prediction models of digital information and grey panel coupling catastrophe prediction models are constructed through the idea of rolling prediction and the method of pattern recognition. They are used to measure the risk and probability, and also forecast and early warn the degree of disaster to the ship besetting in ice accidents of Arctic Sea routes from the macro and micro aspects. Finally, the control strategies and simulation results of ship besetting in ice risk of the Arctic Northeastern Sea routes are completed. Based on the above research, the project would build the theory of digital grey information, and has momentous scientific and practical significance for promoting the information and intelligence of Arctic navigation safety management system.

全球变暖使得北极冰层厚度和冰层覆盖面积逐年减少，北极航道通航有望成为现实，而船舶冰困风险预警是实现北极航行安全的重点内容和难点问题。本项目首先面向通航环境信息资源的数字化，研究北极航道要素的灰信息特征及信息融合技术；其次分别从时间特征、空间位置特征、时空特征三个维度，考虑环境风险因素的时滞性和周期性，建立数字化灰信息关联模型群，研究北极航道环境风险因素与船舶航行状态的映射(互动)关系，开展北极航道船舶冰区航行环境风险因素识别和状态划分；然后利用滚动预测思想和模式识别方法，建立数字化灰信息多变量滚动预测模型和灰色面板耦合灾变预测模型，从系统宏观与因素微观两个角度对北极航道船舶航行冰困事故的的风险测度、形成概率及灾害程度进行预测和预警；最后完成北极东北航道船舶冰困风险控制对策制定及效果仿真。项目研究成果对建立数字化灰信息理论，推动北极航行安全管理系统的信息化和智能化具有重要的科学和现实意义。

本项目面向通航环境信息资源的数字化，首先研究数字化灰信息融合与多维分析、数字化灰信息交通行为与风险机理和数字化灰信息交通流预测模型。具体内容包括定义了数字化灰信息的内涵和外延，描述了灰信息的动态演化过程，给出了基于灰信息演化理论的灰聚类方法；对信息覆盖灰数、多源异构数据、Z数等数字化灰信息，给出了信息覆盖灰数差异信息融合和二元语义数据融合方法，建立了三阶张量的数据融合新机制，同时提出了灰信息覆盖交互关联度模型、二元语义矩阵灰关联度模型、基于Z数灰关联度模型、云-随机森林风险评估模型和灰色综合云混合贴近度的多属性群决策方法等；根据交通流的黏弹性特征，解析了短时交通流灰预测建模机理，同时根据数据驱动模式，建立了面向向量数据流、矩阵数据流及张量数据流等数字化灰信息的系列交通流灰预测模型，初步建立数字化灰信息关联和预测模型体系。然后基于北极航道船舶航行环境风险因素的数字化灰信息特征，以及北极航道环境因素、船舶航行状态与冰困风险之间的互动关系，分别建立了基于离散Z-numbers的北极水域船舶冰困故障树分析模型和基于离散Z-numbers灰关联度的北极航道效益评价模型，对北极航道船舶航行冰困事故的风险测度、形成概率进行预测，并开展北极航道通航效益评价。研究结果表明：目前东北航道船舶发生冰困事故的概率约为1.98%，其中恶劣冰况、GPS故障、通信设备故障以及救援失败等是导致冰困事故发生的主要因素；苏伊士航道的综合效益仍然最高，但东北航道的效益一直在稳步上升，北极航道将有希望成为远洋海运的最佳路线。这些研究成果对建立数字化灰信息建模理论，促进北极船舶航行安全管理的数字化和智能化具有重要的科学和现实意义。