Research On Six-dimensional Force Sensor Calibration System Based On Stewart Platform

With the rapid development of science and technology,six-dimensional force sensors are widely used in various fields.Due to the error factors of design principle and manufacturing,a certain deviation exists between theoretical calculation value and the inputs and outputs of the six-dimensional force sensors.To ensure the correctness and reliability of its use,the six-dimensional force sensor must be calibrated after it is designed and processed.The existing six-dimensional force sensor calibration system generally can't achieve continuous automatic high-precision loading.And the design of a high-precision six-dimensional force sensor calibration system which can realize continuous automatic loading is of great significance to the accurate calibration of six-dimensional force sensor.In this paper,the advantages and disadvantages of the typical six-dimensional force sensor calibration system are compared and analyzed.And a novel six-dimensional force sensor is designed,which is consists of a monocular vision system as a positioning device,a Stewart platform as a basal frame and a servo motor cylinder for force loading.The 3D modeling of the six-dimensional force sensor calibration device is built based on Pro/E.The calibration process and the force and torque loading method in three directions of X,Y and Z are given.The research on positioning device of the six-dimensional force sensor calibration device is completed by the camera calibration,feature point extraction and algorithm development of spatial target location.The camera calibration is achieved by using Camera Calibration Toolbox,and the calibration results are analyzed.The feature points are extracted in MATLAB based on Harris operator.The extraction of the target position coordinates in space is completed by the spatial target location algorithm.And the feasibility of the algorithm is verified by simulation experiments.The position kinematics inverse solution of Stewart platform is analyzed.Based on the Stewart platform prototype and ADLINK motion control card,the motion control programming of Stewart platform is completed in C++ by combining with kinematic inverse control algorithm.And the precise control of movement from the pin to the loading point extracted by a monocular vision system is achieved.The control system of the force loading device is designed based on the functional requirements of the calibration system,and the research on the force / torque servo loading control system is completed.A data acquisition application is developed in C++ based on ADLINK data acquisition card,which is used to achieve the signal acquisition and analysis of single-dimensional force sensor.The self-tuning fuzzy PID controller is designed by using fuzzy toolbox in MATLAB.The simulation model of SIMULINK module is used to simulate the controller,and the PID controller algorithm is optimized.