北极航行船舶航线的多目标优化模型研究

The development and utilization of Arctic shipping routes are of great economic and strategic significance for our country since they will help us to reduce the dependence on shipping in conventional routes and mitigate the risk of energy transport. While due to the complexity of arctic meteorological and oceanographic (MetOcean) conditions, such as Ice, wind, wave and current, etc., in the Arctic, current ship navigation methods and tools cannot fulfil the requirements of safety and economic benefits for Arctic shipping. One important measure to ensure the sustainable Arctic shipping is to research and develop multi-objective dynamic routing optimization system, which is based on the variation of Arctic MetOcean conditions and some reliable ship performance models. However, only a few studies have preliminary investigated the Arctic route optimization problem. Furthermore, there exist large uncertainties in the Arctic ship’s performance and risk models, which are required for routing optimization in Arctic shipping. All these lead to unreliable voyage planning from the Arctic routing optimization. In order to solve these problems, this project will systematically investigate the variation and evolution mechanism of the Arctic MetOcean conditions, build Spatio-Temporal ice, wind and wave crossing-variation models, and implement transform Gaussian/Laplace process to simulate the coupling MetOcean conditions in Arctic routes. Then, big data techniques are introduced to be combined with theoretical analysis, empirical formulas, and experimental tests in ship performance analysis. It improves the modelling accuracy of mathematical models to describe a ship’s safety and energy consumption in Arctic sailing. This project will also build mathematical models to describe a ship’s daily operation cost in Arctic operation. Innovative algorithms will be developed to combine the dynamic programming and genetic algorithm for routing optimization. Finally, the routing optimization dynamic management system will be developed through the integration of all ship performance models, operation cost models and Spatio-temporal MetOcean models by the innovative optimization algorithm for safe and economic efficiency Arctic shipping.

北极航道的开发和利用，对于减少我国对常规航线的依赖、降低能源运输的安全风险，具有重要的经济和战略意义。然而北极航道海冰气象环境复杂多变，当前指导船舶航行的方法和手段，无法满足北极航行对安全性和经济性的要求。研发基于气象环境变化的多目标航线优化决策系统，是实现北极有效航行的重要方向。目前国内外对北极航线的优化算法研究刚刚起步，实现优化决策所需要的船舶性能和风险模型存在很大的不确定性，导致北极航线的规划决策比较盲目。本课题将针对上述问题，系统分析北极气象环境的变化趋势和演变机理，建立海冰况时空耦合特征模型，运用高斯/拉普拉斯随机过程进行模拟；并将大数据机器学习方法，与船舶理论分析、经验公式和实验测试相结合，提高北极航行船舶安全性和能效性建模的准确度；并对船舶北极日常运营成本进行数学建模，研发动态规划和基因算法相结合的多目标优化算法，对上述模型进行集成，形成北极航道船舶航线的动态优化决策系统。

北极航道的开发和利用，对于减少我国对常规航线的依赖、降低能源运输的安全风险，具有重要的经济和战略意义。然而北极航道海冰气象环境复杂多变，当前指导船舶航行的方法和手段，实现优化决策所需要的船舶性能和风险模型存在很大的不确定性，导致北极航行的规划决策比较盲目，无法满足北极航行对安全性和经济性的要求。本项目研发了基于气象环境变化的多目标航线优化系统，其集成了北极航行气象环境的感知、船舶航行操作性能的模拟、气象导航、航行风险决策和规划航行的操作实施等辅助航行决策功能等本课题产出的研究成果，为实现北极船舶的安全有效航行提供了重要的技术保障。本课题按照研究计划已经实施并完成的主要研究内容包括：系统分析了北极气象环境的变化趋势和统计特性，建立了海冰况时空耦合特征模型，运用拉普拉斯随机过程模型对不同的冰况和风速进行了模拟；将数据分析与船舶理论推理和经验公式相结合，提高了北极航行船舶安全性和能效性模型的准确度；并研发动态规划多目标优化算法，从而形成北极航道船舶航线的动态优化决策系统。通过实验和实测数据验证，本课题开发的北极航行决策系统可以至少延长50%的船舶疲劳寿命；根据航行路线和季节的不同，可以节省大概3-8%的燃油消耗。本课题所构建的北极航行风险模型，可以显著降低船舶在北极冰况航行中的冰困和撞击结构损伤等引起的航行风险。开发的北极气象环境随机相关性时空模型，不仅可以用来模拟船舶在北极航行可能遭遇的风浪环境，还可以对北极区域的极限气象风速进行有效地预测，而且可以和其它数据分析方法进行结合，预估来年北极船舶航行的可能时间窗口，这些研究成果已经在不同的北极船舶航行和决策中得到了很好的体现。此外，此课题的研究成果对我们更好地了解北极气象环境的变化、船舶航行的风险演变机理、冰与船的相互作用和响应关系等基础工程和科学问题，开辟了新的方法和思路。