On the Performance of Online Parameter Estimation Algorithms in Systems with Various Identifiability Properties

In recent years, Bayesian inference has been extensively used for parameter estimation in nonlinear systems; in particular it has proved to be very useful for damage detection purposes. The problem of parameter estimation is inherently correlated with the issue of identifiability, i.e., is one able to learn uniquely the parameters of the system from available measurements? The identifiability properties of the system will govern the complexity of the posterior probability density functions (pdfs), and thus the performance of learning algorithms. Off-line methods such as Markov Chain Monte Carlo methods are known to be able to estimate the true posterior pdf, but can be very slow to converge. In this paper we study the performance of on-line estimation algorithms on systems which exhibit challenging identifiability properties, i.e., systems for which all parameters cannot be uniquely identified from the available measurements, leading to complex, possibly multi-modal posterior pdfs. We show that on-line methods are capable of correctly estimating the posterior pdfs of the parameters, even in challenging cases. We also show that a good trade-off can be obtained between computational time and accuracy by correctly selecting the right algorithm for the problem at hand, thus enabling fast estimation and subsequent decision making.

[1]  R. Handel,et al.  Can local particle filters beat the curse of dimensionality , 2013, 1301.6585.

[2]  Rudolph van der Merwe,et al.  The unscented Kalman filter for nonlinear estimation , 2000, Proceedings of the IEEE 2000 Adaptive Systems for Signal Processing, Communications, and Control Symposium (Cat. No.00EX373).

[3]  Jinwoo Jang,et al.  Development of Data Analytics and Modeling Tools for Civil Infrastructure Condition Monitoring Applications , 2016 .

[4]  Andrew W. Smyth,et al.  Review of Nonlinear Filtering for SHM with an Exploration of Novel Higher-Order Kalman Filtering Algorithms for Uncertainty Quantification , 2017 .

[5]  Ian F. C. Smith,et al.  Iterative structural identification framework for evaluation of existing structures , 2016 .

[6]  Olivier Brun,et al.  Parallel Particle Filtering , 2002, J. Parallel Distributed Comput..

[7]  The Split and Merge Unscented , 2009 .

[8]  Simon J. Godsill,et al.  An Overview of Existing Methods and Recent Advances in Sequential Monte Carlo , 2007, Proceedings of the IEEE.

[9]  Arnaud Doucet,et al.  A survey of convergence results on particle filtering methods for practitioners , 2002, IEEE Trans. Signal Process..

[10]  P. Bickel,et al.  Curse-of-dimensionality revisited: Collapse of the particle filter in very large scale systems , 2008, 0805.3034.

[11]  Andrew W. Smyth,et al.  On the observability and identifiability of nonlinear structural and mechanical systems , 2015 .

[12]  Friedrich Faubel,et al.  The Split and Merge Unscented Gaussian Mixture Filter , 2009, IEEE Signal Processing Letters.

[13]  Fredrik Gustafsson,et al.  Marginalized adaptive particle filtering for nonlinear models with unknown time-varying noise parameters , 2013, Autom..

[14]  J. Beck,et al.  Bayesian Updating of Structural Models and Reliability using Markov Chain Monte Carlo Simulation , 2002 .

[15]  Nando de Freitas,et al.  Rao-Blackwellised Particle Filtering for Dynamic Bayesian Networks , 2000, UAI.

[16]  A. Doucet,et al.  A Tutorial on Particle Filtering and Smoothing: Fifteen years later , 2008 .

[17]  P. L. Green.  Bayesian system identification of a nonlinear dynamical system using a novel variant of Simulated Annealing , 2015 .

[18]  J. Ching,et al.  Transitional Markov Chain Monte Carlo Method for Bayesian Model Updating, Model Class Selection, and Model Averaging , 2007 .

[19]  Jeffrey K. Uhlmann,et al.  A consistent, debiased method for converting between polar and Cartesian coordinate systems , 1997 .

[20]  Simo Särkkä,et al.  Bayesian Filtering and Smoothing , 2013, Institute of Mathematical Statistics textbooks.

[21]  Andrew W. Smyth,et al.  Online Noise Identification for Joint State and Parameter Estimation of Nonlinear Systems , 2016 .

[22]  Andrew W. Smyth,et al.  Online Bayesian model assessment using nonlinear filters , 2017 .

[23]  Jeffrey K. Uhlmann,et al.  Reduced sigma point filters for the propagation of means and covariances through nonlinear transformations , 2002, Proceedings of the 2002 American Control Conference (IEEE Cat. No.CH37301).

[24]  Simon J. Julier,et al.  Skewed approach to filtering , 1998, Defense, Security, and Sensing.

[25]  J. Beck,et al.  UPDATING MODELS AND THEIR UNCERTAINTIES. II: MODEL IDENTIFIABILITY. TECHNICAL NOTE , 1998 .

[26]  A. Doucet,et al.  Gradient-free maximum likelihood parameter estimation with particle filters , 2006, 2006 American Control Conference.

[27]  Xiaodong Luo,et al.  Scaled unscented transform Gaussian sum filter: Theory and application , 2010, 1005.2665.

[28]  Nathaniel C. Dubbs,et al.  Comparison and Implementation of Multiple Model Structural Identification Methods , 2015 .

[29]  J. Beck,et al.  Asymptotically Independent Markov Sampling: A New Markov Chain Monte Carlo Scheme for Bayesian Interference , 2013 .

[30]  Simon J. Julier,et al.  The scaled unscented transformation , 2002, Proceedings of the 2002 American Control Conference (IEEE Cat. No.CH37301).

[31]  Andrew W. Smyth,et al.  Particle filtering and marginalization for parameter identification in structural systems , 2017 .

[32]  Fredrik Lindsten,et al.  An efficient stochastic approximation EM algorithm using conditional particle filters , 2013, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[33]  Andrew W. Smyth,et al.  Model updating of a full-scale FE model with nonlinear constraint equations and sensitivity-based cluster analysis for updating parameters , 2017 .

[34]  James L. Beck,et al.  Bayesian Updating and Model Class Selection for Hysteretic Structural Models Using Stochastic Simulation , 2008 .

[35]  Yong Huang,et al.  Approximate Bayesian Computation by Subset Simulation using hierarchical state-space models , 2017 .

[36]  Jeffrey K. Uhlmann,et al.  Unscented filtering and nonlinear estimation , 2004, Proceedings of the IEEE.

[37]  Maria Pia Saccomani,et al.  DAISY: A new software tool to test global identifiability of biological and physiological systems , 2007, Comput. Methods Programs Biomed..

[38]  Thomas B. Schön,et al.  Marginalized particle filters for mixed linear/nonlinear state-space models , 2005, IEEE Transactions on Signal Processing.

[39]  James L. Beck,et al.  Bayesian Updating and Model Class Selection of Deteriorating Hysteretic Structural Models using Seismic Response Data , 2007 .