Research On Shape Adaptive And Nondestructive Grabbing For Marine Biological Machine Fishing

With the expansion of the sea ranch scale,especially the sea cucumbers,scallops and sea urchins,the aquatic fishing have not changed,and the traditional aquatic fishing have become less suitable for the development trend of seafood aquaculture.In order toexplore new seafood fishing methods and realize the mechanization and unmannedization of aquatic fishing,combining the key project of The National Natural Sciense Foundation-"Research on Autonomous Environment Sensing of Underwater Robotics and Nondestructive Grasping of Biological Targets"(item number 70100160005),a new type of underwater-actuated shape adaptive manipulator is used to replace humans with robots as carriers to realize seafood fishing.In this paper,under-actuated four-finger grasping mechanisms are designed for different forms of sea creatures,and the modularization idea is applied to the design of the grasping mechanism,which improves the efficiency of assembly and maintenance of the robot.The tendon rope-torsion spring combined finger transmission method is designed,and the parameter design of the torsion spring at different joints is completed,which improves the envelope grasping effect of underactuated fingers and expands the working space of the fingersThe kinematics of the manipulator was analyzed and a dynamic model was established.At the same time,a static model of the finger was established for the shape diversity of sea cucumber,scallop and sea urchin targets,and the grasping posture of the claw was analyzed for the characteristic size.The relationship between power take-off and joint torque.In order to achieve non-destructive gripping,a six-dimensional torque sensor was designed to measure the gripping force information to achieve the closedness of the gripping force.A finite element analysis was performed on the sensor's elastomer.A calibration experiment was performed on the sensor.The sensor output data was obtained by inputting a standard force.The calibration data was processed using the least square method to obtain a calibration matrix.Finally,the error of the sensor was analyzed.The control link of the manipulator is an important link to complete the grasping task.In the article,the finger kinematics and the static and dynamic models of the previous analysis are used to perform independent PD control simulation and force control simulation of the finger grasp.In this paper,the author previously designed a large-envelope envelope gripping manipulator for the project team,and then carried out grab tests on the underactuated manipulator and large-envelope envelope manipulator designed in this article to analyze the Crawl performance.