Complicated channel geometry and currents may aggravate the interactions between passing ships and berthed ships, which should be evaluated and taken into account in a port design. A method for predicting the ship-shi...Complicated channel geometry and currents may aggravate the interactions between passing ships and berthed ships, which should be evaluated and taken into account in a port design. A method for predicting the ship-ship interactions, based on a non-hydrostatic shallow water flow model, is presented in this paper and is validated by comparing the numerical results with experimental data. The method is subsequently applied to predict the interaction forces acting on a berthed ship due to a passing ship in ports. The influences of the difference of the water depths between the dock and the main channel, the dock geometry, the current and another berthed ship in the dock on the ship-ship interactions are studied. Analysis based on the numerical results is carried out, which is useful for the port design.展开更多
针对风力助航船舶航线优化中存在的风能利用效率量化不足、油耗预测精度受限以及多目标协同优化机制缺失等问题,提出1种融合动态风帆控制与混合驱动预测的多目标航线优化方法。通过建立基于流体力学特性的动态风帆控制策略模型,实现风...针对风力助航船舶航线优化中存在的风能利用效率量化不足、油耗预测精度受限以及多目标协同优化机制缺失等问题,提出1种融合动态风帆控制与混合驱动预测的多目标航线优化方法。通过建立基于流体力学特性的动态风帆控制策略模型,实现风帆辅助推力的空间矢量解析,该模型突破传统静态攻角设定的局限性,可即时动态调整帆角参数,使风能转化效率处于较高水平。为解决传统物理模型环境适应性差与数据驱动方法物理可解释性弱的双重局限,构建物理约束下的人工神经网络分层融合架构,通过船舶运动学方程构建特征空间基底,采用注意力机制引导的人工神经网络进行残差学习。该方法在保留能耗物理机理的同时,实现数据特征与流体力学方程的双向耦合,经北大西洋航线的验证表明,其油耗预测平均绝对百分比误差(mean absolute percentage error,MAPE)较纯物理模型降低21.9%,较纯数据驱动方法的可解释性也大大提升。在此基础上,建立包含时间成本和燃油消耗的多目标优化模型,设计基于非支配排序遗传算法(non-dominated sorting genetic algorithm,NSGA-Ⅱ)和逼近理想解排序法(technique for order preference by similarity to ideal solution,TOPSIS)的协同优化算法,其非劣解集收敛速度较标准算法得以提升。以“新伊敦”轮为对象的实证研究表明:优化后的航线在北大西洋典型航次中,风帆有效工作效率提升,相较于传统推荐航线,优化航线的单航次航行时间缩短5%左右,油耗成本和固定成本分别降低9.1%和4.95%,总成本降低超过7.2%,有效的提高了风力助航船的经济效益并较少了对环境的污染。展开更多
基金Project supported by the National Natural Science Foundation of China(Grant No.51061130548)
文摘Complicated channel geometry and currents may aggravate the interactions between passing ships and berthed ships, which should be evaluated and taken into account in a port design. A method for predicting the ship-ship interactions, based on a non-hydrostatic shallow water flow model, is presented in this paper and is validated by comparing the numerical results with experimental data. The method is subsequently applied to predict the interaction forces acting on a berthed ship due to a passing ship in ports. The influences of the difference of the water depths between the dock and the main channel, the dock geometry, the current and another berthed ship in the dock on the ship-ship interactions are studied. Analysis based on the numerical results is carried out, which is useful for the port design.
文摘星载合成孔径雷达(Synthetic Aperture Radar,SAR)在轨处理技术对提升遥感应用效率至关重要,能解决传统模式下星地数传带宽瓶颈和数据处理延迟问题,在军事侦察、灾害应急监测等领域意义重大。然而,现有在轨处理硬件系统在功能覆盖和处理复杂任务能力方面存在不足,大多针对单个处理环节或简单任务进行研究,无法满足复杂星载任务需求。本文设计了一种大规模星载SAR处理系统,以应对这些挑战。该系统支持数据预处理、SAR成像、舰船检测、重聚焦、目标识别、几何矫正等多种操作,具备广域舰船检测和广域舰船识别两种复杂功能。通过提出基于多节点协同重构的星载SAR多任务兼容方法,详细设计了系统硬件方案,包括交互主控板、检测板、成像板、识别板等硬件板卡。交互主控板负责数据输入、预处理和分发;检测板实现目标检测、重聚焦等功能;成像板利用专用片上系统(System on Chip,SoC)芯片进行高速SAR成像;识别板完成虚警剔除和目标识别。同时,针对广域舰船检测和识别功能,分别设计了相应的数据流和流水线。在广域舰船检测中,数据经预处理后分发至各板卡,成像后进行目标检测、虚警剔除、动目标定位和几何矫正;广域舰船识别在检测基础上增加重聚焦环节,以实现精准识别。实验采用模拟信号源仿真星载SAR回波数据,对系统功能和性能进行验证。结果显示,SAR成像功能的分辨率、峰值旁瓣比和积分旁瓣比均满足要求,目标检测发现率达93.7%,虚警密度为2.96个/万平方公里。在处理时效性方面,广域舰船检测和识别任务分别达到1∶2和1∶2.5准实时处理水平。综上,该系统处理质量良好,成像速度出色,为星载SAR大规模在轨处理技术发展提供了借鉴。
文摘针对风力助航船舶航线优化中存在的风能利用效率量化不足、油耗预测精度受限以及多目标协同优化机制缺失等问题,提出1种融合动态风帆控制与混合驱动预测的多目标航线优化方法。通过建立基于流体力学特性的动态风帆控制策略模型,实现风帆辅助推力的空间矢量解析,该模型突破传统静态攻角设定的局限性,可即时动态调整帆角参数,使风能转化效率处于较高水平。为解决传统物理模型环境适应性差与数据驱动方法物理可解释性弱的双重局限,构建物理约束下的人工神经网络分层融合架构,通过船舶运动学方程构建特征空间基底,采用注意力机制引导的人工神经网络进行残差学习。该方法在保留能耗物理机理的同时,实现数据特征与流体力学方程的双向耦合,经北大西洋航线的验证表明,其油耗预测平均绝对百分比误差(mean absolute percentage error,MAPE)较纯物理模型降低21.9%,较纯数据驱动方法的可解释性也大大提升。在此基础上,建立包含时间成本和燃油消耗的多目标优化模型,设计基于非支配排序遗传算法(non-dominated sorting genetic algorithm,NSGA-Ⅱ)和逼近理想解排序法(technique for order preference by similarity to ideal solution,TOPSIS)的协同优化算法,其非劣解集收敛速度较标准算法得以提升。以“新伊敦”轮为对象的实证研究表明:优化后的航线在北大西洋典型航次中,风帆有效工作效率提升,相较于传统推荐航线,优化航线的单航次航行时间缩短5%左右,油耗成本和固定成本分别降低9.1%和4.95%,总成本降低超过7.2%,有效的提高了风力助航船的经济效益并较少了对环境的污染。
