An a priori map is often unavailable for a mobile robot in a new environment. In a large-scale environment, relying on manual guidance to construct an environment map will result in a huge workload. Hence, an autonomous exploration algorithm is necessary for the mobile robot to complete the exploration actively. This study proposes an autonomous exploration and mapping method based on an incremental caching topology-grid hybrid map (TGHM). Such an algorithm can accomplish the exploration task with high efficiency and high coverage of the established map. The TGHM is a fusion of a topology map, containing the information gain and motion cost for exploration, and a grid map, representing the established map for navigation and localization. At the beginning of one exploration round, the method of candidate target point generation based on geometry rules are applied to extract the candidates quickly. Then, a TGHM is established, and the information gain is evaluated for each candidate topology node on it. Finally, the node with the best evaluation value is selected as the next target point and the topology map is updated after each motion towards it as the end of this round. Simulations and experiments were performed to benchmark the proposed algorithm in robot autonomous exploration and map construction.

中文翻译：

---

基于TGHM算法的移动机器人自主探索与地图构建。

在新环境中，移动机器人通常无法使用先验图。在大规模环境中，依靠手动指导来构建环境图将导致巨大的工作量。因此，对于移动机器人来说，自主探索算法对于主动完成探索是必要的。这项研究提出了一种基于增量缓存拓扑-网格混合映射（TGHM）的自主探索和映射方法。这样的算法可以高效，高覆盖已建立的地图来完成探索任务。TGHM是拓扑图和网格图的融合，其中包含用于勘探的信息增益和运动成本，而网格图则表示已建立的导航和定位图。在一个探索回合的开始，应用基于几何规则的候选目标点生成方法，快速提取候选对象。然后，建立TGHM，并为其上的每个候选拓扑节点评估信息增益。最后，选择具有最佳评估值的节点作为下一个目标点，并在每次回合结束后更新拓扑图。进行了仿真和实验，以验证该算法在机器人自主探索和地图构建中的性能。选择具有最佳评估值的节点作为下一个目标点，并在每次回合结束后更新拓扑图。进行了仿真和实验，以验证该算法在机器人自主探索和地图构建中的性能。选择具有最佳评估值的节点作为下一个目标点，并在每次回合结束后更新拓扑图。进行了仿真和实验，以验证该算法在机器人自主探索和地图构建中的性能。