基于无人水面艇感知网的目标船航迹关联与数据融合

针对现有无人水面艇（Unmanned Surface Vehicle，USV）在航行过程中感知周围航行目标时出现的数据源单一、数据延迟、数据丢失等问题，提出一种基于USV搭载的航海雷达和全球定位系统（GPS）数据源的USV海上航行目标感知数据融合方法。基于最小误差法提出雷达原始图像数据解析算法，并采用数据剔除、时间空间统一方法完成对目标数据预处理，构建基于欧氏距离和马氏距离的航迹关联算法模型、基于层次分析法（Analytic Hierarchy Process，AHP）和专家评价法的融合数据权重分配模型。同时，开展USV试验研究，验证整体融合方法。结果表明，目标原始数据预处理方法合理可靠，融合算法稳定可信，可为USV海上航行目标感知、安全航行及快速避碰提供技术和算法支持。

Target Ship Track Association and Data Fusion Based on Sensing Network of Unmanned Surface Vehicle

In view of the existing problems such as single data source, data delay, and data loss that occur when Unmanned Surface Vehicle (USV) senses the surrounding navigation targets during the navigation, a data fusion method of USV maritime navigation target sensing based on the data source of marine radar and GPS is proposed. The original radar image data analysis algorithm is proposed based on the minimum error method, the target data preprocessing is finished with the data elimination and time-space unification method, and the track association algorithm model based on Euclidean Distance and Mahalanobis Distance and the fusion data weight distribution model based on Analytic Hierarchy Process (AHP) and expert evaluation method are constructed. At the same time, the USV test researches are conducted to verify the integrated fusion method. The results show that the original target data preprocessing method is reasonable and reliable, and the fusion algorithm is stable and believable, which can provide technical and algorithm support for the USV maritime navigation target sensing, safe navigation, and rapid collision avoidance.