Estimation and Prediction of Long-Term Drift of the High Resistance Standards
摘要
The drift of the standards can significantly affect the calibration results of measuring instruments, so such standards require regular calibration using more accurate standards. Taking into account historical data on the drift of the reference standard allows us to assess how its drift can affect the calibration results of other measuring instruments. A requirement is set for reference standards for their stability—the limit of change in the measured value during the intercalibration interval. The article presents the results of assessing the long-term drift of reference resistance standards of direct current (DC) with nominal values of 1 MΩ, 10 MΩ, 100 MΩ and 1000 MΩ using polynomial regression of 1–3 orders. Expanded measurement uncertainties for calibration of DC resistance standards are estimated from 6 to 9 μΩ/Ω. The results of the latest calibrations of these resistance standards are used to predict the time drift for future calibrations. The predicted time drift period for the measurements of 3 years was chosen taking into account their intercalibration interval of 1–2 years. The difference between the obtained predicted values is from 2 to 6 μΩ/Ω from the average values of the standards used. The predicted values of the standards were simulated using the Monte Carlo method for linear and exponential regressions. Those values differ slightly from the obtained data using the Monte Carlo method, but are within the established confidence interval.