Compliance control based on Particle Swarm Optimization approach for physical human-robot interaction

Abstract

Robots play more important roles in daily life and bring us a lot of convenience. Somebody may need to cooperate with robots to complete some specific tasks. But when people work with robots, there remain some significant differences in human-human interactions and human-robot interaction. It is our goal to make robots look even more human-like. We hope that the robot can following the hand of human who act force on the robot. We design a controller which can sense the force acting on any point of the robot and responds indirectly to the force. In addition, a novel design method used to determine the optimal controller parameters of the physical model using the Particle Swarm Optimization (PSO) algorithm is presented. Ultimately, we have carried out experiments on the Spinal Surgery System Robotic (RSSS II), which is developed by our team. The result shows that the new controller performs better in the process of human-robot interaction.