Font Size: a A A

Dynamic Measurement Of Multi-axis Force Sensor Using Errors-in-Variables Model

Posted on:2021-01-27Degree:MasterType:Thesis
Country:ChinaCandidate:M S WangFull Text:PDF
GTID:2428330605952331Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
Multi-axis force sensor is a kind of sensor which can measure the force,pressure or torque in two or more directions.It is widely used in scientific research and industrial production.In some applications,multi-axis force sensor is designed for dynamic measurement.However,at present,most force sensors,which used in dynamic measurement is statically calibrated,and there is no consistency between the measured value and the measurement model.In order to get more reliable measured,appropriate dynamic measurement model should be established and analyzed,and also its measurement uncertainty.Therefore,this paper studies the key technologies of calibration,inversion and measurement uncertainty in the dynamic measurement of multi-axis force sensor.During the dynamic calibration process of force sensor,In order to identify the noise model automatically,the periodic dynamic force is used as the excitation force source,and the non-parametric model is used to describe the noise.The Errors-In-Variables(EIV)frequency domain identification model of force sensor transfer function is established.The parameters of transfer function are obtained by maximum likelihood estimation,and the cost function that is minimized by Levenberg-Marquardt algorithm.MDL criterion is used to determine the order of transfer function.According to the covariance matrix of the parameters of the transfer function,the uncertainty of the frequency response of transfer function is further evaluated.In the inversion of multi-axis force sensor dynamic measurement,a time-domain deconvolution method is used to build the inversion model.In order to solve the ill posed problem of deconvolution,the truncated SVD regularization method is used to approximate the solution.The simulation results show that the method is effective.In the calculation of measurement uncertainty,the frequency-domain method is adopted because the uncertain information of transfer function cannot be fully utilized in the inversion process of time-domain method.Considering the uncertainty of transfer function and the uncertainty of sensor output signal,a dynamic measurement model of multi-axis force sensor is established.The calculation process of measurement uncertainty is analyzed by using Monte Carlo Method(MCM).The results show that the measurement scheme can give the dynamic measured information of multi-axis force sensor completely.In the end,the multi-axis force sensor was used to measure the contact stress between tire and pavement.The preliminary measurement results are given in this paper,and the uncertainty of the measurement indicates that the results are effective and accurate.
Keywords/Search Tags:multi-axis force sensor, dynamic measurement, sensor calibration, errors in variables model, measurement uncertainty
PDF Full Text Request
Related items