Manipulation Task Simulation of a Soft Pneumatic Gripper Using ROS and Gazebo

摘要

Bioinspired robotics have shown great advantages for manipulation tasks performing, compared to current rigid structure with limited degrees of freedom. And the research of pneumatic soft robotic grippers has flourished greatly in recent years. However, there is still no effective platform available for robotic manipulation task simulation when equipped with such a deformable gripper. This paper intends to demonstrate a simple but effective approach of how to create simulator of robotic system with soft gripper and implement robot control in a short time. Here we build this platform based on Robot Operation System (ROS) and Gazebo engine. First, we define an unified module description format for gripper configurations. Then, a dynamic model is formulated to control the action of each joints and gripper deformation after simulated pneumatic actuation. Finally, the entire simulation system including modeling, kinematics, control, and visualization is established. Experimental results from pick&place manipulation of several irregular objects have shown that the proposed simulation platform could achieve an easy configuration, good integration and functional visualization for task simulation of robotic system equipped with a soft pneumatic gripper, and provide a flexible technique for quick verification of prototype or algorithms in the research of soft robotics.