摘要
根据测量仪器校准参数数据特点,建立了仪器校准参数动态发展趋势的时间序列组合模型。将灰色预测与马尔柯夫预测方法优化互补,用灰色预测模型预测参数总体发展趋势,在此基础上采用马尔柯夫模型预测参数在总体趋势下的随机波动性变化,得到校准参数发展趋势预测模型的解。给出基于校准参数趋势预测的校准间隔动态优化方法,并通过实验数据对预测模型进行了验证。结果表明,此模型既能预测校准参数总体趋势,又适合于波动性较大的随机序列变化;校准间隔的优化改善了计量管理中不足计量和过剩计量的缺陷。
According to the character of calibration parameters, a data-preprocessed statistical models for forecasting the calibration trends of measuring instrument is proposed. The solution of the statistical models is got by the merits combination of both gray forecast and Markov forecast, a gray system model is used to forecast the general trend of the calibration status, and a time series Markov chain model is used to forecast data' s fluctuant changing along the general trend. The calibration interval is optimized according to the forecast trends. A preliminary validation of the models is provided based on a collected sample of experimental data. Results demonstrate that the model can well and truly forecast the evolvement and changing trends of the calibration status, the optical calibration interval improved the problem of insufficient calibration and superfluous calibration.
出处
《计量学报》
CSCD
北大核心
2007年第2期184-187,共4页
Acta Metrologica Sinica