Study On Bayes Modeling And Predicting For Dynamic Measurement Errors And Measurement Uncertainty

Nowadays, in order to improve dynamic measurement precision effectively, the correction technique for dynamic errors has become one of the main techniques. At present, there are many methods for dynamic error predicting. However, their predicting precision will decrease when the predicting step increasing. To limit the predicting precision loss in a certain range, author presented a method of Bayes modeling and predicting for dynamic errors based on standard value interpolation at intervals during the multi-step prediction after consulting a lot of papers at home and abroad. The author also gave the uncertainty of the prediction values in this paper according to Bayes theorem. The method can be used to correct the original error model conveniently and effectively.Considering that measurement system uncertainty will change when its working time is extended, Bayes modeling method for the dynamic character of measurement system uncertainty is also discussed in the paper.The above theories are proved to be correct and reliable in the experiments made on a closed loop dynamic measurement system. Besides, to prove whether the above theories are suitable for open loop dynamic measurement system, the author gave some suggestions to improve a grating length measurement instrument that is used for dynamic precision research to make sure that the device could really realize dynamic measurement.