高精度的海上船舶轨迹预测是降低船舶碰撞风险、提升船舶搜救效率的重要基础.海上航行环境的多变性使船舶轨迹数据在时间和空间上具有高度复杂性,现有方法对船舶轨迹数据的质量及运动信息关注度不足,难以充分捕捉轨迹中的时空特征和关联信息.因此,文中提出融合数据质量增强和时空信息编码网络的船舶海上轨迹预测方法(Ship Maritime Trajectory Prediction Method Integrating Data Quality Enhancement and Spatio-Temporal Information Encoding Network,DQE-STIEN).首先,基于船舶轨迹数据的特征,设计结合哈希映射分类及局部离群哈希值异常检测的数据质量增强算法,对问题数据进行质量增强.然后,针对多属性的船舶轨迹数据,设计具有双编码通道的时空信息编码网络,充分提取并整合船舶轨迹数据中的位置信息与运动特征.最后,基于时空信息编码提取数据中的时空关联信息,并经解码生成完整的轨迹预测结果.在5个不同区域的AIS数据集上的实验表明DQE-STIEN性能较优.同时,DQE-STIEN具有一定的泛化性,也能有效分析能源、销售、环境和金融等领域的时序数据.
Constructing an information storage or communication system, where countless pieces of information canbe hidden like a canvas and revealed on demand throughspecific stimuli or decoding rules, is significant. In the presentstudy, we developed a hydrogel canvas that leverages noncovalentinteractions to induce phase separation in the polymer matrix, creating various “paintings”, including custommessages, using different chemical inks. Our strategy focuseson designing small molecule inks, with varying affinities withthe hydrogel and specific responsiveness to stimuli, to achievemultiple changes such as color shifts, fluorescence emission,and dynamic optical image evolution. This skips the typicaldesign approaches, such as incorporating responsive fluorophoresinto polymers for color emission through grafting orcopolymerization, and thus avoids the complex processes involved in modifying and synthesizing functional polymers,along with the uncertainties in material properties that theseprocesses bring.
