Font Size: a A A

Collision Identification And Security Control For Industrial Robots

Posted on:2015-06-23Degree:MasterType:Thesis
Country:ChinaCandidate:G K WuFull Text:PDF
GTID:2308330461474901Subject:Mechanical and electrical engineering
Abstract/Summary:PDF Full Text Request
Industrial robots are mainly used in modern industrial production, they often complete cumbersome and repetitive tasks for human beings. With the development of technology, the application fields of industrial robots are gradually widening. Nowadays besides being widely used in manufacturing field, industrial robots are also used in other fields, such as military, service and so on. However, the premise for robot being widely used in other fields is that robot should be safe enough. The security issue has restricted further development and application of robot technology, and it has become a topic in the field of robotic researches. This study came from National Natural Science Foundation project ’Actively Improving the Safety of Robot-arm by Feedback Control of the Force Generated from the Deformation of Viscoelastic Covering". The main purpose of this paper is to present a collision detection algorithm and a safety control algorithm for robot. Concretely, the collision detection algorithm should be capable of detecting collisions occurred on any position of the robot, and identifying the location and orientation of the impact force. The safety control strategy should ensure the security of the robot.Firstly, this paper established the forward and inverse kinematics of the industrial robot by D-H method, then proposes a method to get proper inverse kinematics solutions. This method is based on the characteristics of the solutions and the geometric structure of the robot. In addition, this paper established the dynamic equations for general serial robot, which is based on the method that combined Screw Theory and Lagrangian Method. Then a concrete form of dynamic equations for the robot in the lab was built, which lays a theoretical foundation for collision detection algorithm and safety control strategy.Secondly, this paper presents a collision detection algorithm based on generalized momentum deviation observer. This observer is designed according to the decoupling performance between generalized momentum and the external torque at robot joints. By using this collision detection algorithm, the equivalent moment of the external force at robot joints can be obtained. When comparing the equivalent moment with torque threshold, we can find whether the robot collides or not. The collision detection algorithm has a good real-time performance and a good accuracy and it can observe which linkage collides. Furthermore, the algorithm can get the direction of the impact force when considering the geometry of robot. In addition, implementation of the algorithm on robot is disscussed. The experiment results show that collision detection algorithm can accurately track the external force and has a good real-time performance.Finally, this paper proposes a safety control algorithm to ensure the safety of robot. This algorithm can ensure robot’s safety no matter at which part of robot the collision happened. The algorithm, which is based on the information provided by the collision detection algorithm, can be applied to industrial robot that usually works only under position or speed control. In addition, this paper discusses a method to achieve safety physical human-robot interaction. A robot motion control software has been programmed by using VC++6.0. Experiments verified the effectiveness of the robot safety control algorithm.
Keywords/Search Tags:industrial robot, collision detection algorithm, safety control, physical human-robot interaction, dynamic model
PDF Full Text Request
Related items