Design Of The Six-dimensional Force Sensor For Dexterous Fingertip

Six-dimensional force sensors have been widely used in engineering practice,especially in the aerospace field.When engaged in precision operations such as onorbit repair,you need to use smart hands to complete some complex operations.Installing six-dimensional force sensors on smart hands can achieve contact Accurate force/torque detection of points to ensure the compliant control of force/torque.Based on the application background and related technical requirements,this paper summarizes the research status of the six-dimensional force sensor sensitive components and the technical problems that still need to be solved in engineering applications.It proposes a new design idea for the six-dimensional force sensor and gives its detection.Principles and decoupling algorithms.Combined with the research and technical background of elastomers at home and abroad,comprehensively considering the size and force of the dexterous hand,the external dimensions and technical indicators of the six-dimensional force sensor are clarified,and an improved cross beam planar elastomer structure is proposed.Combining the cross beam and the force transmission beam,the patch on the force transmission beam measures the horizontal force,thereby reducing the number of patches on the plane of the elastomer,and making its structure more compact,ensuring the miniaturized structure design of the dimensional force sensor,and the ANSYS finite element analysis software is used to perform stress analysis on the elastomer,determine the paste position of the resistance strain gauge,and design the external package.According to the different full-scale output of each channel,a corresponding Wheatstone bridge is built to convert the external load into strain and output an electrical signal.A two-stage amplifier circuit is designed to amplify the output of the bridge and collect the voltage output.In order to certain the performance of the six-dimensional force sensor,simulate the calibration experiment,use ANSYS to measure the strain output of different loads,use Multisim to design the simulation circuit,add environmental variables such as zero drift and temperature drift,and maximize the simulation of the real environment.Establish an algorithm model for the coupling error between dimensions,use the least square method and BP neural network algorithm to decouple the collected data,and compare the advantages and disadvantages of the two algorithms and the decoupled calibration error and Dimensional coupling error.