Control of grip force with fuzzy logic for precision grasp task of robotic hand

In this study, it is aimed to ensuring control of the grasping force in order to grip and lift an unknown object without permanent deformation and dropping in the precision grasp task of a robotic hand. Grasping objects is a complex task involving uncertainties for the robotic hand. Humans are able to perform this task successfully. The fuzzy logic method was used for ensuring control of the grip force by transferring human abilities oriented to grasping object to machine. When humans grip and lift objects, they add extra force for safety above the minimum limit value of the grasping force. This extra force is expressed in the literature as the "Safety margin". The percentage ratio of safety margin according to minimum limit value of grasping force is expressed as "Safety margin ratio". In the first step of this study, precision grip-lift experiments were conducted with humans. Optimum safety margin data were obtained from these experiments. As a result, it was found that humans apply safety margin between 9% and 20% by varying ratios, depending on the varying surface friction property and weight. A fuzzy safety margin calculation system was designed by using the data obtained from the experiments. In the performance tests, the fuzzy safety margin calculating system calculated safety margin for the same object conditions in the experiments between 8% and 21% by varying ratios. Later, the actuator which representing the fingers of a robot hand was designed and manufactured. The design of the grip force control system was realized by integrating the fuzzy safety margin calculation system, actuator and other control and command elements. In the designed system, the local weight approach was developed, which allows the safety margin to be more optimized. In the literature was stated that the safety margin ratio varies between 10% and 40% and in some studies was accepted as a constant %20. The grip force control system designed in this study was gripped and lifted objects in real-time by the safety margin varying between 3% and 16% depending on changing object conditions. The obtained results were demonstrated the applicability of the method and approach in this study for the control of grip force in the precision grasp task of the robot hand.