Roadmap composition for multi-arm systems path planning - LAAS - Laboratoire d'Analyse et d'Architecture des Systèmes Accéder directement au contenu
Communication Dans Un Congrès Année : 2009

Roadmap composition for multi-arm systems path planning

Résumé

This paper presents a new method for planning motions of multi-arm systems in constrained workspaces, for which state-of-the-art planners behave poorly. The method is based on the decomposition of the system into parts. Compact roadmaps are first computed for each part, and then, a super-graph is constructed by the composition of elementary roadmaps. Results presented for a three-arm system and a model of the complex DLR's Justin robot show a significant performance gain of such a two-stage roadmap construction method with respect to single-stage methods applied to the whole system. I. INTRODUCTION Multi-arm robot systems have been developed in diverse fields such as industrial manufacturing [1], medical robotics [2], and humanoid robotics [3]. Such complex high-dof systems may have to perform tasks in constrained workspaces, in which computing feasible paths is a very difficult task. Robot motion planning has been an active research domain over the past decades [4]. The path planning problem consists in finding a feasible path between two given configurations of a mobile system. Feasible paths have to satisfy intrinsic constraints of the system (e.g. mechanical design constraints, kinematic constraints), as well as constraints that arise from the environment (e.g. collision avoidance). Using the notion of configuration space [5], the problem is reduced to explore the connectivity of the subset of the feasible configurations. Sampling-based planners are able to solve complex problems in high-dimensional spaces with very low computational cost. One of the most popular sampling-based planners is the Probabilistic RoadMap (PRM), introduced in [6] and further developed in many other works (see [7], [8] for a survey). The PRM algorithm has been shown to perform well for a broad class of problems. However, its performance degrades in the presence of narrow passages, which require prohibitively dense roadmaps in order to be captured. A number of variants and extensions have been proposed to alleviate this problem, e.g. biasing sampling around obstacles [9], [10], [11] or towards the medial axis [12], using free-space dilatation [13], [14], visibility-based filtering [15] exploiting search space information [16], or delaying collision detection [17], [18]. Despite the established efficiency of PRM-like planners, the construction of a roadmap enabling to solve constrained problems for multi-arm systems is very expensive because of the high-dimensionality of the configuration-space (typically, around 20 DOF for a torso with two arms), which is
Fichier principal
Vignette du fichier
09iros.pdf (658.28 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01986305 , version 1 (18-01-2019)

Identifiants

  • HAL Id : hal-01986305 , version 1

Citer

Mokhtar Gharbi, Juan Cortés, Thierry Simeon. Roadmap composition for multi-arm systems path planning. 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2009), Oct 2009, St. Louis, United States. pp.2471-2476. ⟨hal-01986305⟩
21 Consultations
12 Téléchargements

Partager

Gmail Facebook X LinkedIn More