This research presents the kinematics analysis and the teleoperation system of a 4-DOF modular reconfigurable robot. The forward kinematic analysis of the designed robot is conducted firstly. A numerical method with the backpropagation neural network is investigated to solve inverse kinematics problem and the training samples for the neural network are obtained with FARO laser tracker, which guarantees modeling accuracy of the non-linear mapping from the task space to the joint space. The remote control system has a client-server structure. The local robot control computer plays a role as a server and remote terminal as a client. Simulation, monitoring and control of the robot can be conducted in a remote manner. When client works in the simulation mode, users can simulate robot motion with the virtual one. While the system works in the monitoring/control regime, the robot can be remotely controlled and the status of robot can be monitored dynamically at client site. By using Java3D technique, virtual model of the robot is implemented and only small data parcels with the current robot coordinates need to be transmitted to the visualization module of client side. This accelerates the system response and provides the operator with a synthesized view of the real robot.
Keywords: Kinematics analysis, teleoperation, reconfigurable robot, modular robot, java3d, remote control, forward kinematic, inverse kinematics, neural network, visualization module.
