船舶航迹预测是其避障、导航的基础，精准的航迹预测可提升船舶在复杂水域航行的安全性，为此研究基于高阶常微分方程的复杂水域船舶航迹精准预测方法。该方法利用AIS系统采集船舶航行行为动态观测样本后，构建该船舶航行行为动态观测样本的高阶常微分方程预测模型，使用四阶龙格-库塔方法求解该预测模型，得到船舶航行动态观测样本预测值后，利用最小二乘法对其进行拟合处理，得到船舶航迹曲线。实验结果表明，该方法可有效采集船舶在复杂水域航行时的航速、船首向等行为动态观测样本，预测船舶航迹点和拟合后的航迹曲线均与其实际数值吻合，预测精度较高。
Ship track prediction is the basis of obstacle avoidance and navigation, and accurate track prediction can improve the safety of ship navigation in complex waters. Therefore, the accurate prediction method of ship track in complex waters based on higher order ordinary differential equations is studied. This method uses the AIS system to collect the dynamic observation samples of the ship's navigation behavior, and then constructs the high order ordinary differential equation prediction model of the dynamic observation samples of the ship's navigation behavior. The fourth order Runge-Kutta method is used to solve the prediction model. After obtaining the predicted values of the dynamic observation samples of the ship's navigation behavior, the least squares method is used to fit them to obtain the ship's track curve. The experimental results show that this method can effectively collect the dynamic observation samples of ship's speed, bow direction and other behaviors when the ship is sailing in complex waters. At the same time, the predicted ship's track points and the fitted track curves are consistent with the actual values, and the prediction accuracy is high.
2023,45(4): 59-62 收稿日期：2022-10-30
DOI：10.3404/j.issn.1672-7649.2023.04.012
分类号：TP391.9
基金项目：江西省教育厅科技项目(GJJ181567)；国家自然科学基金资助项目(12163003)
作者简介：陈云龙(1977-),男,硕士,讲师,从事微分方程数值解研究
