Neuro-fuzzy in six DOF tele-robotic control

Abstract

A force-force bilateral scheme based on neuro-fuzzy control was designed for a six DOF tele-robotic system. An open architecture controller for the six DOF tele-robotic system has been successfully implemented. Increased system bandwidth system can be achieved with the new embedded PUMA 760 and PUMA 260 controllers. Both the slave and master robot controllers comprise a PC running a real-time operating system, a multi-axis motion controller, and a PCI force/torque data acquisition unit. Furthermore, the complex soft computing (master and slave kinematics, force acquisition, and neuro-fuzzy) was implemented using multi-tasking concepts over a real-time local network. With the new system design, and inter-processing master-slave controller computation can significantly improve the bandwidth of the system, and both the master and slave robot subsystems can be operated with high robustness of stability and manoeuvrability, and easier to fully implement and control.