Slip Detection Based Reflex Control Of A Prosthetic Hand

In unstructured environments, although the material, weight, stiffness of the grasped object are unknown and external disturbances are uncertain, human also can adjust the grasping force according to the contact information on the hand appropriately, realizing stable grasp without slip. However, for a prosthetic hand, which is lack of humanoid flexibility and efficient skin receptors, it is a serious challenge to make a stable manipulation in unstructured environments. Especially, when the applied force is too small, how to deal with the dynamic external disturbances and avoid slips becomes an urgent scientific problem that must be solved. To reconstruct the humanoid reflex control function on a prosthetic hand, we proceeded from slip detection and reflex control scheme, and did further investigation on reflex control with non-slip. The main works were as fellows.(1) On the basis of analyzing the construction characteristics of the tendon-driven prosthetic hand, Dynamical equations were derived with the method of Lagrange equation. Analytical model and3D prototype were simulated in MATLAB and ADAMS respectively, confirming the reasonableness of the model.(2) A new-type slip sensor was designed using a force sensing resistor and techniques of STFT was applied to do frequency spectrum analysis of the output signals. A new method that using the Empirical Mode Decomposition was brought up to detect slippages. The performance of the slip detection scheme was tested through experiments study and the results verified the effectiveness and accuracy of the scheme.(3) Considering the changes of the stiffness and slip speeds of the grasped objects, we built a reflex force mode based on the fuzzy logic control. The humanoid reflex control scheme was brought up and an adaptive PI controller based on sliding speed was designed. Simulations were done with different objects to test the feasibility and correctness of the reflex control strategy.(4) Based on the reflex control experimental platform, grasp control experiments were carried out with different objects that got different slip speeds. The experimental results showed that the prosthetic hand reflex control system we designed had a good performance in force self-adaptive adjustment, effectively preventing objects slip, which verifies the validity of the entire control system.