DSL-Assembly: A Robust and Safe Assembly Strategy
share
A reinforcement learning (RL) based method that enables the robot to accomplish the assembly-type task with safety regulations is proposed. The overall strategy consists of grasping and assembly, and this paper mainly considers the assembly strategy. Force feedback is used instead of visual feedback to perceive the shape and direction of the hole in this paper. Furthermore, multiple models based on different sensors are trained for different environments due to environmental perturbations and equipment failures (failures of cameras and other sensors) in the real world. Then, since the emergency stop is triggered when the force output by the robot is too large, a force-based dynamic safety lock (DSL) is proposed to limit the pressing force of the robot. Finally, we train and test the robot model with a simulator and build ablation experiments to illustrate the effectiveness of our method. The models are independently tested 500 times in the simulator, giving a 58.91 world and deployed on a real robot. Simulation environments: https://github.com/0707yiliu/peg-in-hole-with-RL.
READ FULL TEXT