基于窄脉冲校准的高量程加速度传感器动态建模  

Dynamic Modeling of High Range Acceleration Sensor Based on Narrow Pulse Calibration

作　　者：米振国 石云波[1] 王彦林 温晓杰 张娟娟[1] MI Zhenguo;SHI Yunbo;WANG Yanlin;WEN Xiaojie;ZHANG Juanjuan(Science and Technology on Electronic Test&Measurement Laboratory,North University of China,Taiyuan Shanxi 030051,China)

机构地区：[1]中北大学电子测试技术重点实验室,山西太原030051

出　　处：《传感技术学报》2020年第8期1155-1160,共6页Chinese Journal of Sensors and Actuators

基　　金：电子测试重点实验室稳定支持经费项目(WD61420010411804)。

摘　　要：面对高量程加速度传感器动态校准中存在的测不准以及数据处理困难等问题,论述了窄脉冲校准原理,引入了消除噪声干扰的广义最小二乘算法对于传感器的校准数据进行处理,并且结合窄脉冲校准原理进行了算法的数值仿真与实验验证。利用Hopkinson杆对于自研的量程为200 000 g_n的多级封装传感器进行了窄脉冲动态校准实验,系统辨识的结果表明模型获得的传感器的本征模态与通过FFT方法获得的本征模态分别为482.4 kHz以及481.8 kHz,并且曲线精度大大优于后者。Faced with the problems of inaccuracy and difficulty in data processing in the dynamic calibration of high-range acceleration sensors,the principle of narrow pulse calibration is discussed,and the generalized least squares algorithm to eliminate noise interference is introduced to process the calibration data of the sensor,and combining with the narrow pulse calibration principle has carried on the numerical simulation of the algorithm and experimental validation. The narrow-pulse dynamic calibration experiment for the self-developed multistage package sensor with a range of 200 000 g_n was carried out by Hopkinson lever. The system identification results showed that the intrinsic modes of the sensor obtained by the model and those obtained by FFT method were 482.4 kHz and 481.8 kHz respectively,and the curve accuracy was much better than the latter.

关 键 词：高量程加速度传感器 动态校准 系统辨识 广义最小二乘法 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]