Dynamic Obstacle-Avoidance Human-Robot Collaboration System for Industry Robots
Abstract: Nowadays, human-robot collaboration technologies are realized in controlling different typesof robots in industry. Even now the current robot technology is very advanced, due to therobot's freedom space and complex operating environment, which makes robot pathplanning issues more complicated. Human-robot collaboration is the trend for industrialfactories in the future,as it both provides the accuracy of robots and the flexibility of thehuman. In this case, combining robot technology with computer virtual control and simulationplatforms makes it possible to carry out research on robots without a real robot.Obstacle avoidance of a robot arm is an important factor to be considered in the pathplanning for industry production. When there is any obstacle, like if human appears in theoperating environment, real-time detection of obstacles can effectively prevent the roboticarm from colliding with the obstacle.Based on a literature review about current path planning algorithms and the applied situation,this master thesis concentrates on developing a system with an improved path planningmethod called T-RRT(Transition-based Rapidly-exploring Random Tree) to realize that thissystem can prevent the robot from the obstacle collision in real-time. This thesis provides abasic prototype about dynamic avoidance in a simulation environment both with the respectto the obstacles and the Universal Robot 5. The results show that the T-RRT algorithm caneffectively plan a new path when there is any obstruction in the ongoing path. Besides, thedetection secure distance represented by a bounding box will change along with the speedof the obstacle.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)