DBSCAN算法邻域参数在碎片分类任务中的应用及效果

As shown in Figure 10, the neighborhood parameters in the DBSCAN algorithm are specifically represented in the task context as 'velocity increment limit' and 'fragment density within a hypersphere'. The six sets of neighborhood parameters for the tasks are (2km/s, 4 fragments), (2km/s, 3 fragments), (1.5km/s, 3 fragments), (3km/s, 4 fragments), (3km/s, 5 fragments), and (3km/s, 6 fragments). The transfer time between fragments is set to 100s for the spacecraft to perform clustering, resulting in 45, 118, 58, 139, 96, and 67 fragments being selected and divided into 14, 41, 22, 45, 30, and 22 fragment clusters, respectively.