Multi-objective optimisation for humanoid robot motion planning. (2016)

• Record Type: Journal Article
• Authors: Xu, Xiandong; Guan, Yi; Hong, Bingrong; Ke, Wende; Zhong, Qiubo; Piao, Songhao
• Is Part Of: International journal of wireless and mobile computing. Volume 10:Number 2(2016)
• Page Start: 112
• Page End: 121
• Publication Date: 2016
• Subjects: humanoid robots -- NSGA-II -- genetic algorithms -- self-adjusting binary search trees -- multi-objective optimisation -- robot motion -- motion planning -- robot gait -- walking robots -- legged locomotion -- gait optimisation -- parametric control -- slope walking -- gait planning; Mobile computing -- Periodicals; Wireless communication systems -- Periodicals; 004.6
• Languages: English
• ISSNs: 1741-1084

Abstract

Single objective optimisation method is primarily adopted to solve humanoid robot gait optimisation problems. However, this method has certain limitations because it can consider only one objective function, such as energy, stability, or speed. This study presents a method based on the improved Non-Dominated Sorting Genetic Algorithm-II (NSGA-II) and parametric control technique to optimise the parameters of the humanoid robot when walking up and down a slope. We propose a novel non-dominated sorting method based on a self-adjusting binary search tree, which overcomes the low efficiency of the traditional fast non-dominated sorting. We use the improved NSGA-II with this new non-dominated sorting method to achieve multi-objective optimisation of the gait parameters for a humanoid robot walking on a slope. Experiment results indicated that this method is effective and can better realise gait planning for humanoid robots walking on a slope.