离心机法校准加速度计不确定度的统计模拟算法

加速度计精密离心机法校准的复杂性,使误差源的不确定度传播系数不满足显式解析、可导和近似线性的条件,因而无法应用常规的微积分方法求解。统计模拟方法是复杂测量不确定度计算的有效方法。通过模拟产生转速和半径随机矢量的多次实现值,经数据变换和最小二乘拟合得到加速度计方程系数的样本空间,最后用统计方法求得系数的不确定度,与离心机实验法校准结果比较,具有较高的一致性,证明了统计模拟方法的有效性。

Statistical simulation algorithm to analyzing the uncertainty in calibration of accelerometers by centrifuge

Because of the complication of accelerometer＇s calibration by centrifuge,the uncertainty＇s propagation coefficients dissatisfy the condition of being an explicit,analytical and approximate linear expression.So it＇s impossible to solve this problem with conventional calculus method.Statistical simulation method is effective in calculating the uncertainty of complicated measurement.By simulating the vector values of the centrifuge＇s rotational speed and working radius many times,and then by data conversion and least squares fitting,we got a sample space of the accelerometer＇s equation coefficients,and finally got the uncertainties of the coefficients with statistical method.The result was in a good accordance with that by centrifuge experiment method,and this shows the effectiveness of statistical simulation method.