Bayesian optimal sensor placement for parameter estimation under modeling and input uncertainties

A Bayesian optimal sensor placement (OSP) framework for parameter estimation in nonlinear structural dynamics models is proposed, based on maximizing a utility function built from appropriate measures of information contained the input–output response time history data. The gain quantified using Kullback–Leibler divergence (KL-div) between prior and posterior distribution model parameters. design variables may include type location sensors. Asymptotic approximations, valid large number data, provide valuable insight into measure information. Robustness to uncertainties nuisance (non-updatable) parameters associated with modeling excitation considered by expected over all possible values In particular, handles case where measured installed sensors but remains unknown at experimental phase. Introducing stochastic models, taken uncertain used random variability input histories. Monte Carlo or sparse grid methods estimate multidimensional probability integrals arising formulation. Heuristic algorithms are solve optimization problem. effectiveness method demonstrated multi-degree freedom (DOF) spring-mass chain system restoring elements that exhibit hysteretic nonlinearities.