Design Of Galfenol Cantilever Three Dimensional Force Tactile Sensor For Intelligent Manipulator

In recent years,the research and application of intelligent manipulator have been widely concerned and become an indispensable part of industrial production and daily life.In order to ensure the manipulator to complete the predetermined tasks reliably and deftly,it is necessary to equip the intelligent manipulator with sensors with fingertip information perception ability.Tactile sensing technology is the basis and support of autonomous control and intelligent decision of intelligent manipulator,and force perception is an important research content of perception.Based on the summary of previous studies,this paper developed a new three-dimensional force tactile sensor suitable for the intelligent manipulator by using Galfenol alloy as the core sensing material and combining theoretical modeling with experimental analysis.Firstly,the sensor output characteristics of Galfenol alloy cantilever beam were studied.Based on Euler-Bernoulli beam,nonlinear energy equalization model and hall effect,the sensing output model of iron gallium alloy cantilever beam was derived.The experimental platform was built,the influence of bias magnetic field and cantilever beam size on the sensor output characteristics was investigated,the optimal excitation magnetic field was determined,and the accuracy of the ferrogallium alloy cantilever beam sensing output model was verified,which was the theoretical basis for further optimization of magnetostrictive cantilever beam sensing.Then,according to the output characteristics of Galfenol cantilever beam sensor,the Galfenol cantilevered three dimensional force tactile sensor was designed.Using the Comsol simulation platform,the bidirectional coupling simulation model of the three dimensional force sensor is established,the effectiveness of excitation mode is analyzed,and the feasibility of the three dimensional force sensor is expounded.The step response and dynamic response of the sensing unit are tested and the frequency of the self-damped harmonic and the relationship between the dynamic harmonic and dynamic frequency are determined.In order to reduce the self-damped harmonic interference and the dynamic harmonic interference,the butterworth low pass filtering algorithm is studied.And combined with the needs of the manipulator dynamic sensing,using the Matlab signal processing toolbox in the filter analysis design tool,the design of a three dimensional force sensor digital filtering system.Based on levenberg-marquardt algorithm,the decoupling method of BP neural network is proposed.Finally,the static calibration of the Galfenol cantilevered three dimensional force tactile sensor was carried out and the sample data of the training neural network was obtained.The BP neural network including the input layer,the hidden layer and the output layer was built by using the neural network toolbox of Matlab software,and the training of the network was completed by using the sample data.The filtering and decoupling algorithm is embedded into the STM32 digital processing system to complete the design of the three dimensional force sensor information acquisition and processing module.The experiment verifies the feasibility of the filtering algorithm and decoupling algorithm of the sensor,and the grasping test is carried out with the manipulator.