Bayesian Fusion Method in Different Overall Distribution Between Prior Data and Field Data.

Abstract: Bayesian method is good to improve precision and earn cost, but it would be constrain to use in situation of field data be different with prior data in overall distribution, as generally change model or give up Bayesian method in this situation. In this paper, with consideration of physical connection of prior data and field data, thoughts of using Bayesian fusion method in different overall distribution between prior data and field data was developed, which was from 3 conditions of function of prior data and field data: distribution form and parameters were known, parameter had prior distribution and only distribution form were know. At last, an example was given to explain the usage method promoted by this paper. [Copyright &y& Elsevier]