Font Size: a A A

Research On Network Control For Internet-based Robot

Posted on:2005-11-24Degree:MasterType:Thesis
Country:ChinaCandidate:C J ZhaoFull Text:PDF
GTID:2168360122975657Subject:Pattern Recognition and Intelligent Systems
Abstract/Summary:PDF Full Text Request
With the development of computer network, teleoperations from any place at any time in the world become possible by utilizing it. Internet-based robots will be applied to hazardous or uncertain environments such as remote medical treatment, outer space or industrial manufacture. The main contribution of this paper is to present an architecture of robot network control system, and develop a 3D visual system by using OpenGL and Visual C++6.0. Furthermore, we are engaged in researches on method of robot network control with variational time-delay units and obstacle avoidance algorithm.At first, we discuss the research status and problems in Internet-based robots. Secondly a predictive control method based on RBF neural network was presented to deal with the random time-delay. Furthermore, an avoidance algorithm based on the C space was designed to enhance the robot's intelligence. At last we develop a visual model of PUMA560 robot with OpenGL and a platform for visual simulation. A data exchange protocol plays an important role in communication between two places. It redounds to exchange more types data in only one communication channel. In communication function the Winsock interface was adopted to coding so that the software can be run on more operating systems. This paper proposes a VR-based telerobotic system for providing both visual and rapid information. The proposed system reduces the influence of time-delay on real time.This paper has presented in-depth research on Internet-based robots system, and a visual simulation system of network control was implemented. Experimental results show the efficacy of this approach. It will play an active role in development of robotics along the information and AI way.Research partially supported under the Jiangsu Natural Science Funds(BJ98057).
Keywords/Search Tags:Robot, Network Control, Virtual Reality, Neural Network, Obstacle Avoidance
PDF Full Text Request
Related items