一种基于聚类分组的快速联合兼容SLAM数据关联算法

针对在移动机器人同时定位与建图（SLAM）过程中如何快速准确获取数据关联结果的问题，提出了一种基于DBSCAN（density-based spatial clustering of application with noise）聚类分组的快速联合兼容SLAM数据关联算法（DFJCBB）．首先，采用局部关联策略将参与关联的特征点限定在局部地图中；其次，针对多数环境中量测都有较明显的分布，采用一种基于密度聚类的方法DBSCAN对当前时刻的量测进行分组，从而得到若干关联度小的观测小组；最后，在每个小组中采用联合兼容分支定界（JCBB）算法进行数据关联，以获得每个小组量测与局部地图特征之间的最优关联解，并将这些关联解组合获得最终的关联结果．基于模拟器和标准数据集的仿真实验验证了该关联算法的性能，结果表明该关联算法在保证获得较高关联准确度的同时，大大降低了算法复杂度、缩短了运行时间，适用于解决不同复杂环境中的SLAM数据关联问题．

A Fast Joint Compatibility Data Association Algorithm for SLAM Based on Clustering

In order to quickly and accurately obtain data association results in simultaneous localization and mapping (SLAM) of mobile robot, a fast joint compatibility data association algorithm (DFJCBB) for SLAM based on DBSCAN (density-based spatial clustering of application with noise) is proposed. Firstly, the local association strategy is used to limit features in local map. Then, a density-based clustering method, that is DBSCAN method, is used to group all measurements at the current moment and get a number of measurement groups with small correlation, because the measurements appear a clear distribution in most environments. Finally, joint compatibility branch and bound (JCBB) algorithm is adopted in data association of each group to obtain the optimal association solution between each group of measurements and local map features, and the optimal association solutions are combined to obtain the final association result. The performance of the proposed algorithm is verified by simulation based on the simulator and benchmark dataset. The results show that the proposed algorithm can guarantee high association accuracy, reduce the computational complexity and shorten the running time. It is suitable for solving the data association problem of SLAM in different complex environments.