Multi-agent localization from noisy relative pose measurements

Abstract

Abstract-In this paper we address the problem of estimating the poses of a team of agents when they do not share any common reference frame. Each agent is capable of measuring the relative position and orientation of its neighboring agents, however these measurements are not exact but they are corrupted with noises. The goal is to compute the pose of each agent relative to the anchor node from noisy relative pose measurements. We present an strategy where, first of all, the agents compute their orientations relative to an anchor node. After that, they update the relative position measurements according to these orientations, to finally compute their positions. As contribution we discuss the proposed strategy, that has the interesting property that can be executed in a distributed fashion. The distributed implementation allows each agent to recover its pose using exclusively local information and local interactions with its neighbors. Besides, it only requires each node to maintain an estimate of its own orientation and position. Thus, the memory load of the algorithm is low compared to methods where each agent must also estimate the positions and orientations of any other agent.

A convergence analysis for pose graph optimization via Gauss-Newton methods

“…For sake of simplicity we tailor the formulation to the single robot batch SLAM problem. However, the results also apply to the case in which the different poses in the pose-graph correspond to different agents: this is the case in which the pose graph models multi robot localization [14] or multi robot SLAM [15] problems.…”

Distributed Consensus on Robot Networks for Dynamically Merging Feature-Based Maps

“…This common frame needs to be computed at least once, and usually only requires the robots to know the relative pose of its nearby teammates, see e.g., [29]- [31] where different methods for computing robot-to-robot measurements are presented. There exist several distributed algorithms that combine these measurements to produce the common frame, e.g., [32]- [34], or our recent work [35] and references therein.…”

Distributed trajectory estimation with privacy and communication constraints: A two-stage distributed Gauss-Seidel approach

“…(16) highlights that DGS only requires the estimates for poses involved in its inter-robot measurements E α S . The approach involves almost no "privacy violation": robot δ only sends an estimate of its rendezvous poses.…”

“…Franceschelli and Gasparri [15] propose a gossipbased algorithm for distributed pose estimation and investigate its convergence in a noiseless setup. Aragues et al [16] use a distributed Jacobi approach to estimate a set of 2D poses. Knuth and Barooah [17] estimate 3D poses using distributed gradient descent.…”