Skill Learning for Robotic Insertion Based on One-shot Demonstration and Reinforcement Learning

• 作 者: Ying Li De Xu
• 机 构: Research Center of Precision Sensing and Control,Institute of Automation,Chinese Academy of Sciences
• 出 处: 国际自动化与计算杂志（英文版）
• 发表日期: 2021年3期
• 关键词: Force Jacobian matrix one-shot demonstration dynamic exploration strategy insert

In this paper,an efficient skill learning framework is proposed for robotic insertion,based on one-shot demonstration and reinforcement learning.First,the robot action is composed of two parts:expert action ahd refinement action.A force Jacobian matrix is calibrated with only one demonstration,based on which stable and safe expert action can be generated.The deep deterministic policy gradients (DDPG) method is employed to learn the refinement action,which aims to improve the assembly efficiency.Second,an epis-ode-step exploration strategy is developed,which uses the expert action as a benchmark and adjusts the exploration intensity dynamic-ally.A safety-efficiency reward function is designed for the compliant insertion.Third,to improve the adaptability with different com-ponents,a skill saving and selection mechanism is proposed.Several typical components are used to train the skill models.And the trained models and force Jacobian matrices are saved in a skill pool.Given a new component,the most appropriate model is selected from the skill pool according to the force Jacobian matrix and directly used to accomplish insertion tasks.Fourth,a simulation environ-ment is established under the guidance of the force Jacobian matrix,which avoids tedious training process on real robotic systems.Simu-lation and experiments are conducted to validate the effectiveness of the proposed methods.