Authors: Joong-Tae Park, Jae-Bok Song
Department:Department of Mechanical Engineering, Korea University, Anam-ro 145, Seongbuk-gu, Seoul, South Korea(机械工程系，高丽大学，韩国)
Exploration is one of the most important functions for a mobile service robot because a map is required to carry out various tasks. A suitable strategy is needed to efficiently explore an environment and to build an accurate map. This study proposed the use of several gains (information, driving, localization) that if considered during exploration, can simultaneously improve the efficiency of the exploration process and quality of the resulting map. Considering the information and driving gains reduces behavior that leads a robot to explore a previously visited place, and thus the exploration distance is reduced. In addition, the robot can select a favorable path for localization by considering the localization gain during exploration, and the robot can estimate its pose more robustly than other methods that do not consider localizability during exploration. This proposed exploration method was verified by various experiments, which verified that a robot can build an accurate map fully autonomously and effciently in various home environments using the proposed method.(探索是移动服务机器人最重要的功能之一，因为地图是执行不同任务的必需品。因此需要合适的策略来有效地探索一个环境并且生成准确的地图。这个研究提出如果在探索中使用多个增量[信息，驾驶，定位]，将能同步地提高探索过程的有效性和生成地图的质量。考虑信息和驾驶增量将降低机器人探索之前已经到过的地方的行为，因此探索距离被减小的。除此之外，在探索时考虑定位机器人将能选择对定位有效的路径，因而机器人将比不考虑定位估计姿态更有力。该建议的探索方法由不同的实验验证，结果表明使用该建议的方法机器人能在不同的家庭环境中完全自动而有效地建立准确地图。)
1. Introduction
In recent years, there have been various trials to extend robotic technology to non-industrial applications such as surgery, cleaning, patrol, and so on. Indoor mobile home service robots are receiving attention especially, because of their economic potential and social expectations. In order to make mobile service robots more accessible in home environments, the problem of environmental modeling, which is one of the fundamental problems in mobile robots, should be solved first. This is because a mobile service robot uses a map to carry out various tasks, including navigation, human-robot interaction, and so on. Therefore, the simultaneous localization and mapping (SLAM) community has developed many efficient and highly accurate map-building techniques, but most of these techniques offer no proposals on how a robot can be made to function autonomously. However, autonomy is an important factor for environmentalmodeling of service robots. Therefore, various methods of exploration – the name typically given to automated map-building – have been proposed.
Frontier-based exploration, which explores the unknown area in a grid map, was proposed in. In frontier-based exploration, a robot detects the regions between the unexplored area and the open space, designated as the frontier. The robot then moves to the new frontiers to explore them until the entire environment has been explored. Frontier-based exploration has the draw-back of not being able to use information known about obstacles,which can serve as a guide for the robot to move and correct its localization error. To overcome this problem, an autonomous exploration method using regions of interest was proposed.In this research, the view that would result in the sensor data that could be used to maximize exploration efficiency was estimated. While this approach does improve exploration efficiency, it does not address map accuracy at all. (基于前沿的探索，就是在格网地图上搜索未知区域，被提出来。在基于前沿的探索中，机器人探测未探索的区域和公开空间之间的区域，这部分被划分为前沿。机器人于是移动到新的前沿探索它们直到整个环境都被探索了。基于前沿的探索有不能使用障碍物的信息的缺点，这其实可以作为机器人移动的引导并且纠正定位误差。为了克服这个问题，使用感兴趣区域的自动探索被提出来。在该研究中，可能会导致被用来最大化探索效率的传感器数据的视图得到估计。虽然该方法确实提高了探索效率，但是它并没有解决地图精度。)
The aforementioned strategies are considered metric-based exploration methods. Another type of exploration method exists, known as topological information-based exploration. The most representative topological information-based exploration strategy is based on the Generalized Voronoi Graph (GVG) representation.In Topological SLAM, developed for exploration of an unknown environment, the robot traces all GVG edges and visits all meet points and boundary points
（上述策略被称为基于度量的探索方法。还有另一种探索方法类型的存在，即基于拓扑信息的探索。最具代表性的基于拓扑信息的探索方法是基于广义Voronoi图的表示。）