| Recent decade has witnessed the development and application of dynamic measurement technology, so, the theory of measurement accuracy is confronted with many unknown subjects that need to pursue and solve. Among them, Key theories and technology of analyzing dynamic measurement system, building models of dynamic system accuracy, separating and correcting dynamic measurement errors and evaluating dynamic measurement results, etc appear very clearly in expert's eyes, which are need to be solved immediately.This paper tries to apply Bayesian statistical principle to study accuracy applicable to dynamic measurement, and apply Bayesian dynamic model to research real-time correction technology of dynamic accuracy.Basic steps of building models of dynamic accuracy are given. By inserting standard signals or Bayes one- step forecast, parameters of dynamic accuracy models are adjusted timely to adapt to time-varied property of unsmooth dynamic accuracy and increase prediction accuracy of models. From dynamic data characteristics of ergodic process, uncertainty evaluation of ergodic smooth process is given out with applying Bayesian constant mean value dynamic models in this paper. By Bayesian statistical deduction and prediction, uncertainty of ergodic process is expressed using estimated value of standard deviation of observational equation.The emphases of this paper puts on intervenor of model,groping for a new way and applying it.Finally,design a set of device,give blue print to validate the preciseness of Bayes model. |
