Research On Key Technologies Of Low-frequency Three-component Linear Vibration Calibration Based On Machine Visio

The low-frequency tri-axial vibration sensors are widely applied in seismic monitoring,building structure health monitoring,aerospace navigating and other fields.Their sensitivity calibration accuracy directly determines the validity and reliability of their applications.Although the laser interferometry has been recommended by the ISO to achieve the vibration calibration,it suffers from the shortage of capability in lowfrequency calibration,high cost,low calibration efficiency and limited calibration environment.Meanwhile,as the most commonly used sensor calibration device,the low frequency linear shaker is capable of providing input excitation for vibration sensors in different frequency ranges.However,the dynamic performance of the shaker inevitably changes during the manufacturing,assembling and long-term use of the shaker,which affects the calibration accuracy of the vibration sensor and increases the uncertainty of the sensor calibration.Traditional methods often use laser interferometry and multi-axis sensor method to measure and calibrate its parameters.However,the former has the above-mentioned unavoidable problems in the low-frequency range,and the latter usually has a low signal-to-noise ratio in the low-frequency range,both of which cannot achieve accurate measurements in the low-frequency range.To address the above problems,a three-component linear vibration measurement and calibration method based on machine vision are investigated in this study:(1)A novel monocular vision-based dynamic method is investigated,which determines the spatial sensitivities of tri-axial sensors by the monocular vision method to measure the spatial excitation.This method improves the calibration accuracy by eliminating the installation error and calibration efficiency by decreasing reinstallations.Comparison experimental results demonstrate that the investigated method have the similar calibration results with the laser interferometry in the low frequency range.(2)A binocular vision-based spatial motion measurement method is proposed for the motion characteristics of lowfrequency linear shakers,which is combined with the Zernike moment method to extract the characteristic target motion edges and achieve accurate measurements of dynamic parameters such as amplitude-frequency characteristics,transverse vibration ratio,total harmonic distortion,and repeatability of the shaker.The comparison experiments with laser interferometry and sensor method show that they have similar measurement results,and the machine vision-based method has better low-frequency dynamic range and repeatability.