The use of high-power industrial equipment, such as large-scale mixingequipment or a hydrocyclone for separation of particles in liquid suspension,demands careful monitoring to ensure correct operation. The task of monitoringthe liquid suspension can be posed as a time-evolving inverse problem andsolved with Bayesian statistical methods. In this paper, we extend Bayesianmethods to incorporate statistical models for the error that is incurred in thenumerical solution of the physical governing equations. This enables fulluncertainty quantification within a principled computation-precision trade-off,in contrast to the over-confident inferences that are obtained when numericalerror is ignored. The method is cast with a sequential Monte Carlo frameworkand an optimised implementation is provided in Python.
展开▼
