Theory And Application Of Modern Uncertainty

As an important component of modern metrology and enterprise quality management,measurement uncertainty is an important reference for the objectivity and effectiveness of experimental data.With the wide application of the precision engineering and dynamic testing technology,the theory of measurement uncertainty faces many problems that need to be solved.Especially the error tracing and modeling of the high-precision,multidimensional and complex static measurement systems,the uncertainty modeling and quantification,the uncertainty evaluation of dynamic testing systems and other relevant problems are all the hot issues of the modern uncertainty theory.However,the existing research about the modern uncertainty theory has not yet formed a theoretical system with universal guiding significance,and still been confined to solving the uncertainty evaluation of specific measurement tasks,which either lacks theoretical support or ignores the application.In this paper,the key theories and application difficulties of modern uncertainty are systematically studied.The main work and innovation are summarized as follows:Based on error traceability theory and value statistical analysis,the two methods of uncertainty analysis,the three kinds of model are put forward,which are the measurement model,the analysis model of measurement uncertainty and the evaluation model of measurement uncertainty respectively.The theoretical system for the uncertainty evaluation modeling is set up.The problem of source analysis of the standard uncertainty is solved systematically.The structure form of the universal uncertainty analysis model is given,and the uncertainty evaluation process based on the three models are provided.The problem of the conceptual confusion on the measurement model and uncertainty model has been solved effectively,which has laid a good foundation for the application of MCM.The modern uncertainty evaluation method suitable for dynamic measurement and expert judgment based on Bayesian statistics has been researched.The Bayesian uncertainty optimization and evaluation models have been put forward based on the principle of the maximum entropy after the systematic study of the Bayesian uncertainty evaluation method of non-informative priors and conjugate priority.Based on the theory of the dynamic stochastic process and the principle and model for Bayesian dynamic prediction,the problems of evaluating the dynamic uncertainty of the ergodic random process are studied,the model for uncertainty prediction is established,and the methods for studying the dynamic uncertainty of non-ergodic random process are discussed.The application of MCM in uncertainty evaluation is extended.Based on the Monte Carlo method for evaluating the modern uncertainty suitable for computer simulation and combined with the error combination theory,the uncertainty algebraic sum synthetic method suitable for massive distribution and spreading is researched,which can make the uncertainty evaluation results more in line with the characteristics of error propagation and effectively solve the limitations to the input correlation,as well as the model nonlinearity and complexity in the traditional synthetic method.The method for verifying the modern uncertainty based on the advantages of the MCM random sampling in the uncertainty evaluation is proposed.Organically combined MCM with the Bayesian statistical inference,dynamic uncertainty evaluation,prediction and verification,the modern uncertainty evaluation system for complex measurement systems is established to provide reliable theoretical basis for evaluating and predicting dynamic uncertainty.The task oriented uncertainty evaluation of CMM is studied comprehensively.The uncertainty evaluation models targeting at the dimensions and geometric error measurement tasks of CMM are established.The method for quantization and optimization evaluation of the task-oriented uncertainty components is provided,and the problem of excessive uncertainty estimation is solved.The task oriented uncertainty evaluation software for CMM is programmed,and the automatic identification of the measurement uncertainty intelligent evaluation and optimization evaluation are realized.In this paper,the concept of the uncertainty of CMM space is proposed for the first time,and a three-dimensional uncertainty evaluation model is established,and the identification and quantization method of uncertainty component is given.Moreover,combined with virtual instrument technology and computer technology,the VCMM with measurement accuracy and visual measurement function is designed based on the error transfer mechanism,which provides a new idea for improve the application value of precision instrument.