Nonlinear Error Compensation for Sensor of Stylus Profilometer
The nonlinear error caused by rotation of sensor's lever is an important factor affecting the measuring accuracy of stylus profilometer.Most of existing nonlinear error compensation methods use polynomial to fit and compensate.With the increase of measuring range of the profilometer,the polynomial error compensation methods have larger residual error.Therefore,a non-polynomial error compensation method is proposed for the profilometer.Firstly,a non-polynomial error model is established for sensor of the profilometer.Then,the standard balls are used as standard,Pratt circle fitting least square method and Levenberg-Marquardt optimization algorithm are combined to calibrate the parameters of the model.Finally,the nonlinear error of sensor of the profilometer is compensated by calibrated parameters.The measurement data error of profile of a steel ball with radius of 4.762 5 mm is compensated.The results show that the proposed non-polynomial error model can reduce the nonlinear error of sensor of the profilometer more effectively than traditional polynomial model.In addition,the calibration and compensation experiments of standard balls with different radii show that the calibration parameters obtained by same standard ball can be applied to nonlinear error compensation of curve measurement with different curvature radii,proving that the compensation method can be used to compensate the nonlinear error of sensor during measurement of complex profiles.
rolling bearingprofilometersensornonlinear errorerror compensationleast square method
NETL
NSTL
万方数据
