A Common Optimization Framework for Multi-Robot Exploration and Coverage in 3D Environments
- PDF / 3,118,450 Bytes
- 16 Pages / 595.224 x 790.955 pts Page_size
- 50 Downloads / 173 Views
A Common Optimization Framework for Multi-Robot Exploration and Coverage in 3D Environments Alessandro Renzaglia1,2
· Jilles Dibangoye2 · Vincent Le Doze2 · Olivier Simonin2
Received: 16 December 2019 / Accepted: 4 September 2020 © Springer Nature B.V. 2020
Abstract This paper studies the problems of static coverage and autonomous exploration of unknown three-dimensional environments with a team of cooperating aerial vehicles. Although these tasks are usually considered separately in the literature, we propose a common framework where both problems are formulated as the maximization of online acquired information via the definition of single-robot optimization functions, which differs only slightly in the two cases to take into account the static and dynamic nature of coverage and exploration respectively. A common derivative-free approach based on a stochastic approximation of these functions and their successive optimization is proposed, resulting in a fast and decentralized solution. The locality of this methodology limits however this solution to have local optimality guarantees and specific additional layers are proposed for the two problems to improve the final performance. Specifically, a Voronoi-based initialization step is added for the coverage problem and a combination with a frontier-based approach is proposed for the exploration case. The resulting algorithms are finally tested in simulations and compared with possible alternatives. Keywords Multi-Robot systems · Cooperative exploration · Optimal coverage
1 Introduction Multi-robot teams, especially when involving aerial vehicles, are extremely efficient systems to help humans in acquiring information on large and complex areas [10, 30]. In these scenarios, two fundamental tasks are static coverage and exploration. In the coverage problem, the robots have to find the static set of positions that optimizes a certain coverage criterion, e.g. the portion of the monitored environment. In the exploration, the objective is to generate the paths that Alessandro Renzaglia
[email protected] Jilles Dibangoye [email protected] Vincent Le Doze [email protected] Olivier Simonin [email protected] 1
INRIA, University of Grenoble Alpes, 38000, Grenoble, France
2
INSA Lyon, CITI Lab, INRIA, CHROMA, Lyon, France
allow the robots to observe the entire environment in a minimum time. In both cases, the region of interest is assumed to be unknown and the robots have to retrieve the necessary information during the mission (Fig. 1). For their relevance, these problems have received wide attention in the robotics community and numerous solutions have been proposed in the last years. However, they have been usually treated separately, proposing different formulations and approaches, and not as particular cases of a more general informationbased problem. In both cases, the robots are indeed called to navigate through an unknown environment and cooperate to maximize the observed area. The main contributions of this paper are: to prese
Data Loading...
