A hierarchical optimization control method for whole-body control and arm manipulation of a single-arm wheel-legged robot

作者：Tianshuai Guo, Yaxian Xin, Yu Wang, Xuewen Rong, Baoshuai Sun · 发表于：Journal of Engineering Research · 年份：2026 · DOI：10.1016/j.jer.2026.01.015 · 研究领域：Robotic Locomotion and Control、Prosthetics and Rehabilitation Robotics、Robot Manipulation and Learning

The Single-Arm Wheel-Legged Robot(SAWR) demonstrates the potential for energy-efficient mobility and adaptability to complex terrain, along with operational capabilities. However, manipulator motion disrupts the robot’s dynamic stability,leading to increased balance control complexity and failure in operational tasks. This study introduces a hierarchical optimization control framework designed to guarantee the dynamic stability of a single-arm wheel-legged robot during whole-body motion and arm manipulation tasks.The hierarchical optimization control framework consists of three parts: the arm-torso interaction force estimation layer, the torso control layer, and the force allocation & balance layer. Firstly, the arm-torso interaction force estimation layer computes the generalized forces generated by the arm’s motion and payload in real time and feeds them forward to the torso control layer for disturbance compensation. Secondly, the torso control layer generates the desired generalized forces for the left and right wheel-legs to achieve precise torso pose control with disturbance compensation.The force allocation & balance layer then generates joint torques by converting the desired generalized forces into precise leg force output and dynamic balance posture control. The proposed control framework was validated in Webots simulation environment, showing a marked improvement in the robot’s operational stability. • Single-arm wheel-legged robot with high-speed mobility and terr...