Cooperative Localization in Unknown Environment towards Multi-Robot Systems

• Published in: Chinese Control Conference pp. 4549 - 4554
• Main Authors: Yang, Xiaohang, Zhou, Lelai, Zhang, Chen, Tian, Xincheng, Li, Yibin
• Format: Conference Proceeding
• Language: English
• Published: Technical Committee on Control Theory, Chinese Association of Automation 28.07.2024
• Subjects: Accuracy; Cameras; Cooperative Localization; Cubature Kalman Filtering; GPS-denied environment; Location awareness; Multi-Robot Systems; Prediction algorithms; Relative Localization; Robot vision systems; Robustness; Wheels
• ISSN: 1934-1768
• DOI: 10.23919/CCC63176.2024.10661940

Localization plays an important role in robotics, aiming at the problem of multi-robot cooperative localization in a GPS-denied environment, a multi-sensor fusion cooperative localization method based on relative observations is presented in this paper. The relative observation module is comprised primarily of a monocular camera and an ultra-wideband (UWB), which associates visual detection with UWB distance measurement data to obtain relative measurement models for multi-robot systems. An Extended Cubature Kalman Filter (ECKF) is designed to fuse observations and internal sensor odometry data to improve system accuracy and robustness. The algorithm deploys on a single robot, with a fleet of sharing relative observations. To evaluate the performance of the proposed method, we conduct the experiment in the Gazebo simulation environment and the cooperative localization experiment is carried out on three unmanned robots in the real world. Comparing the global camera position as the ground truth, the results show the accuracy and robustness of our system.