An Overview of Multi-task Control for Redundant Robot Based on Quadratic Programming

Qingkai Li,Yanbo Pang,Wenhan Cai,Yushi Wang,Qing Li,Mingguo Zhao
2023-01-01
Abstract:Redundant degrees of freedom provide a feasible solution for robots to better interact with environments, but it also leads to an increase in the complexity of kinematic and dynamic control. This paper presents an overview of methods that realize the whole-body control (WBC) of a redundant robot with a quadratic programming (QP) approach. We first induce the general QP form from the single-task control method for redundant robots and describe WBC as a multi-task control problem in the form of multi-objective optimization (MOO). Then different QP algorithms for multi-task control with fixed and transitional priority are introduced. Weighted quadratic programming (WQP) and hierarchical quadratic programming (HQP) for fixed priority, and generalized hierarchical control (GHC) and recursive hierarchical projection (RHP) for transitional priority will be explicated, respectively. Finally, a toolkit implementing the above algorithms is provided, based on which simulation results are presented and remarks on different algorithms are given.
What problem does this paper attempt to address?