ADD-RRV for motion planning in complex environments. Issue 1 (14th January 2022)

• Record Type: Journal Article
• Authors: Cai, Peng; Yue, Xiaokui; Zhang, Hongwen
• Is Part Of: Robotica. Volume 40:Issue 1(2022)
• Page Start: 136
• Page End: 153
• Publication Date: 2022-01-14
• Subjects: Motion planning -- Complex environments -- Local space identification -- Principal component analysis -- Bridge test; Robots -- Periodicals; 629.89205
• DOI: 10.1017/S0263574721000436
• Languages: English
• ISSNs: 0263-5747

Abstract

Abstract: In this paper, we present a novel sampling-based motion planning method in various complex environments, especially with narrow passages. We use online the results of the planner in the ADD-RRT framework to identify the types of the local configuration space based on the principal component analysis (PCA). The identification result is then used to accelerate the expansion similar to RRV around obstacles and through narrow passages. We also propose a modified bridge test to identify the entrance of a narrow passage and boost samples inside it. We have compared our method with known motion planners in several scenarios through simulations. Our method shows the best performance across all the tested planners in the tested scenarios.