Simulation Study on a Method to Localize Four Mobile Robots Based on Triangular Formation

摘要

Localization of mobile robots is crucial to the deployment of robots in real-world applications. With infrastructure-based technologies, such as global positioning system (GPS) and indoor positioning system (IPS), robot localization is relatively straightforward. However, the extra cost to maintain the infrastructure can be problematic for certain applications. Other methods, such as dead reckoning and simultaneous localization and mapping (SLAM), have also been actively studied, despite the fact that they do not employ these infrastructure-based technologies. However, in such methods, the accumulated errors are still detrimental. Moreover, the SLAM method is hardly used in pattern-less environments, such as in totally open space. In this paper, we propose a localization method for mobile robots based on cooperative formation. The system does not need positioning infrastructures or existing maps. The concept of the proposed localization method is similar to that of localization methods, which are based on infrastructures: first, three mobile robots build a triangular formation, then, they are fixed at their locations, and the robot team accomplish their beacon-as task, as if it were an infrastructure-based localization system. The last robot is a moving robot, as it moves to the next vertex while localizing itself. The positioning error of the proposed system is much lower than the errors caused by dead reckoning or SLAM algorithms without loop matching method, because the error of the proposed system is not continuously accumulated over time. The performance of the proposed system is compared to the performance of the odometry-based localization method and the SLAM algorithm. Simulations demonstrate that the localization method proposed by us produces less positioning errors than the odometry-based method and the SLAM algorithm.