Hybrid position/force control of a dexterous hand based on fuzzy control strategy

Abstract

The application of fuzzy control strategy to the hybrid position/force control problem of a dexterous hand has been demonstrated. From the measured positions, forces and velocities during a manipulative task, an ordinary fuzzy controller was utilized to control the hand in reaction and motion level. Then the results were compared with conventional PID controllers. Due to simplicity of the fuzzy controller, conventional controllers behave much better than the fuzzy one. Furthermore we have demonstrated that by choosing a dynamical sensitive range of error phase plane (distance vs. velocity), the fuzzy controller behaves far better than the PID one.