Configuration-aware model predictive motion planning for Tractor–Trailer Mobile Robot. (4th March 2023)

• Record Type: Journal Article
• Authors: Ito, Nobuaki; Okuda, Hiroyuki; Suzuki, Tatsuya
• Is Part Of: Advanced robotics. Volume 37:Number 5(2023)
• Page Start: 329
• Page End: 343
• Publication Date: 2023-03-04
• Subjects: Collision avoidance -- motion planning -- model predictive control -- non-holonomic -- Farkas' lemma; Robotics -- Periodicals; Robotics -- Japan -- Periodicals; Robotics; Japan; Periodicals; 629.89205
• DOI: 10.1080/01691864.2022.2126733
• Languages: English
• ISSNs: 0169-1864

Abstract

ABSTRACT: This study proposes a motion planning method with strict obstacle avoidance constraints among polygonal-shaped objects for a tractor–trailer mobile robot. The conventional method generally expresses collision-free constraints by approximating a robot as one circle or several circles or ellipsoids. However, accomplishing moving tasks with minimum margins is desirable, with improved efficiency of space availability when multiple robots operate in one area such as a factory, restaurant, office, or warehouse. The target of this study is a tractor–trailer mobile robot that can transport more goods efficiently, while significantly reducing margins during movement. To control the target, multiple convex polygonal objects expressing the mobile robot and surroundings are transformed into linear simultaneous inequalities using 'Farkas' lemma.' Embedding inequalities into model predictive control yields a motion planner that allows tractor–trailer mobile robot movement and possesses non-holonomic constraints. The proposed model is validated through simulation and real machine experiments. GRAPHICAL ABSTRACT: UF0001