抄録
In this paper, an idea on Markov chain Monte Carlo (MCMC) methodology is applied to inverse analysis for a defect profile identification arising in nondestructive test. A inverse analysis is formulated as a reconstruction of parameters related to material damages. A directed acyclic graph is represented by a linear system for inspection procedures and by a statistical model associated with unknown parameters and testing source inputs. A method for estimating the material parameter is proposed based on the Gibbs sampler. Results of simulation experiments to demonstrate the efficacy of the proposed algorithm are reported.
