Collision-Free Motion Generation in Dynamic Environments with Joint-Space Quadratic Optimization

Purpose – The purpose of this paper is to compare the performance of the robot arm motion generated by neural controllers in simulated and real robot experiments. Design/methodology/approach – The arm motion generation is formulated as an optimization problem. The neural controllers generate the robot arm motion in dynamic environments optimizing three different objective functions; minimum execution time, minimum distance and minimum acceleration. In addition, the robot motion generation in the presence of obstacles is also considered. Findings – The robot is able to adapt its arm motion generation based on the specific task, reaching the goal position in simulated and experimental tests. The same neural controller can be employed to generate the robot motion for a wide range of initial and goal positions. Research limitations/implications – The motion generated yield good results in both simulation and experimental environments. Practical implications – The robot motion is generated based on three different objective functions that are simultaneously optimized. Therefore, the humanoid robot can perform a wide range of tasks in real-life environments, by selecting the appropriate motion. Originality/value – A new method for adaptive arm motion generation of a mobile humanoid robot operating in dynamic human and industrial environments.

Download Full-text

Motion generation for walking exoskeleton robot using multiple dynamic movement primitives sequences combined with reinforcement learning

Robotica ◽

10.1017/s0263574721001934 ◽

2022 ◽

pp. 1-16

Author(s):

Peng Zhang ◽

Junxia Zhang

Keyword(s):

Reinforcement Learning ◽

Degrees Of Freedom ◽

Stability Margin ◽

Joint Space ◽

Motion Generation ◽

Movement Primitives ◽

Dynamic Movement ◽

Exoskeleton Robot ◽

Flexion Extension ◽

Dynamic Movement Primitives

Abstract In order to assist patients with lower limb disabilities in normal walking, a new trajectory learning scheme of limb exoskeleton robot based on dynamic movement primitives (DMP) combined with reinforcement learning (RL) was proposed. The developed exoskeleton robot has six degrees of freedom (DOFs). The hip and knee of each artificial leg can provide two electric-powered DOFs for flexion/extension. And two passive-installed DOFs of the ankle were used to achieve the motion of inversion/eversion and plantarflexion/dorsiflexion. The five-point segmented gait planning strategy is proposed to generate gait trajectories. The gait Zero Moment Point stability margin is used as a parameter to construct a stability criteria to ensure the stability of human-exoskeleton system. Based on the segmented gait trajectory planning formation strategy, the multiple-DMP sequences were proposed to model the generation trajectories. Meanwhile, in order to eliminate the effect of uncertainties in joint space, the RL was adopted to learn the trajectories. The experiment demonstrated that the proposed scheme can effectively remove interferences and uncertainties.

Download Full-text