An Enhanced Bayesian Based Model Validation Method for Dynamic Systems

Abstract

Validation of computational models with multiple correlated functional responses requires the consideration of multivariate data correlation, uncertainty quantification and propagation, and objective robust metrics. This paper presents an enhanced Bayesian based model validation method together with probabilistic principal component analysis (PPCA) to address these critical issues. The PPCA is employed to handle multivariate correlation and to reduce the dimension of the multivariate functional responses. The Bayesian interval hypothesis testing is used to quantitatively assess the quality of a multivariate dynamic system. The differences between the test data and computer-aided engineering (CAE) results are extracted for dimension reduction through PPCA, and then Bayesian interval hypothesis testing is performed on the reduced difference data to assess the model validity. In addition, physics-based threshold is defined and transformed to the PPCA space for Bayesian interval hypothesis testing. This new approach resolves some critical drawbacks of the previous methods and adds some desirable properties of a model validation metric for dynamic systems, such as symmetry. Several sets of analytical examples and a dynamic system with multiple functional responses are used to demonstrate this new approach.