Fully Distributed Joint Localization and Target Tracking With Mobile Robot Networks
Fully Distributed Joint Localization and Target Tracking With Mobile Robot Networks
复制标题
DOI：
10.1109/tcst.2020.2991126
复制
发表时间：
2021-07-01
影响因子：
4.8
通讯作者：
Ren, Wei
中科院分区：
文献类型：
--
作者：
Zhu, Pengxiang;Ren, Wei
本文研究了基于移动机器人网络的关节定位和目标跟踪问题。在这里，一组配备了车载传感器的移动机器人同时定位自己并跟踪多个目标。我们介绍了一种完全分布式的算法，该算法适用于一般的机器人运动、目标过程和测量模型，并且对时变的感知和通信拓扑以及变化的盲目机器人(不直接感知目标的机器人)具有鲁棒性。我们没有将定位和目标跟踪视为两个独立的问题，而是显式地考虑了其中一个问题对另一个问题的影响，并利用它来提高完全分布式环境下的性能。给出了两种新的分布估计。通过使用它们，每个机器人可以仅使用其本地信息和来自其一跳通信邻居的信息来估计自身的姿态(位置和方向)(定位)和目标的状态(跟踪)，同时保持一致性。此外，还证明了对于线性时变模型，在感知和通信图以及系统可观测性非常温和的条件下，估计误差在均方意义下是有界的。通过蒙特卡罗模拟和使用真实世界数据集进行的实验，我们的方法的有效性得到了广泛的证明。实验还表明，联合估计机器人的姿态和目标的状态时，机器人的姿态估计获得了更好的性能。
In this article, we study the problem of joint localization and target tracking using a mobile robot network. Here, a team of mobile robots equipped with onboard sensors simultaneously localize themselves and track multiple targets. We introduce a fully distributed algorithm that is applicable to generic robot motion, target process, and measurement models and is robust to time-varying sensing and communication topologies and changing blind robots (the robots not directly sensing the targets). Instead of treating localization and target tracking as two separate problems, we explicitly account for the influence of one to the other and exploit it to improve performance in a fully distributed context. Two novel kinds of distributed estimates are derived. By employing them, each robot can estimate the pose (position and orientation) of itself (localization) and the states of targets (tracking) using only its local information and information from its one-hop communicating neighbors while preserving consistency. Furthermore, it is proven that, in the case of linear time-varying models, the estimation errors are bounded in the mean-square sense under very mild conditions on the sensing and communication graph and system observability. The effectiveness of our approach is demonstrated extensively through Monte Carlo simulations, and experiments carried out using a real-world data set. It is also shown better performance in the pose estimates of the robots is achieved when jointly estimating the robots' poses and targets' states.