Design And Control Of Underactuated Bionic Manipulator Based On Hand Motion Recognition

The hand is the organ that human beings rely on most in the labor process,and the manipulator is also an extremely important end effector for robots.Theref ore,researchers have designed many different types of manipulators for different usage scenarios and usage requirements.Among them,the underactuated manipulator is adaptive and can cooperate with the palm to grasp the target objects of different shapes.Therefore,under different working conditions,it does not need to replace the end effector,but only needs to modify the control algorithm to complete various forms of However,when the specifications of the target objects are different,frequent changes to the algorithm will cause a lot of waste of time and labor.Based on this,this paper is devoted to combining the existing hand motion recognition technology with the manipulator,and designs an underactuated manipulator that can realize teleoperation to realize adaptive grasping of objects with different shapes.First,the structural characteristics of the human hand are analyzed to determine d the various indicators of the manipulator design.Through the analysis and comparison of the twisted wire transmission mode,the transmission principle is analyzed and the corresponding kinematics with mechanical models are constructed,and the key parameters of the transmission mode are obtained.Based on the transmission principle,the under-actuated structure of the finger module is designed.Then the required motor model is selected based on the design index of the manipulator and the parameters of the twisted wire transmission,and the design ideas of the palm and the thumb are briefly introduced.Secondly,the theoretical basis of manipulator kinematics is introduced,and on this basis,a kinematic coordinate system is established for the manipulator fingers designed in this paper.The pose of each phalanx relative to the reference coordinate system is obtained by operation,and when the pose of the far phalanx is known,the pose corresponding to the proximal and middle phalanges is solved,and the Jacobian matrix of the finger is obtained.Then,the static model of the manipulator in the envelope grasping process is solved.And the two key components of the pin and the proximal knuckle are analyzed by finite-element analysis.Thirdly,the design idea and composition of the manipulator control system are described,and the implementation of the motor drive circuit and the position signal acquisition circuit is introduced.Then,the design idea of the force sensor is analyzed in detail,and the temperature compensation algorithm is designed.Then determine the collection scheme of hand motion,and select the force-position hybrid control algorithm through comparative analysis to process the force-position signal collected by the encoder disc and force sensor.Finally,the effectiveness of the manipulator is verified and analyzed from the experimental point of view.Experiments show that the parameters of the manipulator meet the design requirements,and the manipulator can be remotely operated through the data glove to complete various types of grasping tasks designed.