关键词: 校准周期, 改进的灰色模型, 等维新息, 残差修正, 马尔可夫过程, 预测精度

Abstract:

To solve the determination problem of calibration interval for test instruments in Automatic Test System, we present an improved grey GM (1,1) model forecasting method. First we improve the model itself based on GM(1,1) model, and then build a best equaldimension and newinformation grey model which can realize the best dynamic prediction of original data. In order to further improve the prediction accuracy, we propose a residual error correction model, and Markov process is applied to deal with the symbol problem of residual error correction. Real cases show that the improved model forecasting method has better prediction accuracy and can be applied to the prediction process of calibration interval.

Key words: calibration interval, the improved grey model, equaldimension and newinformation, residual error modification, Markov process, prediction accuracy