Suction control of a robotic gripper: A neuro-fuzzy approach

This paper discusses a fuzzy logic control system designed to determine, re gulate and maintain the amount of suction needed by a robotic gripper syste m to perform reliable limp material manipulation. A neuro-fuzzy approach is followed to determine the amount of desired suction (depending on experime ntally derived data and plant characteristics). A knowledge-based valve con troller is then designed to generate, regulate and maintain the amount of s uction calculated by the neuro-fuzzy suction module. The performance of the overall suction control system is compared with actual experimental result s obtained when using a prototype gripper system to handle limp material. F urther, performance of the fuzzy logic based valve controller is compared t o conventional PD and PID controllers. The proposed control scheme is found to enhance the overall functionality of the prototype robotic gripper syst em.