Model updating of a nonlinear experimental vehicle using substructuring and unscented Kalman filtering
Chatzi, Eleni N.
- Conference Paper
This study establishes a computational framework for nonlinear finite element (FE) model updating of large-scale structures, through the exploration of two case studies pertaining to a spring-mass-damper chain model and to a laboratory vehicle with nonlinear suspensions. The proposed approach combines a substructuring model reduction approach with a near realtime system identification scheme, namely the Unscented Kalman Filter (UKF). The former aims at isolating and locally updating individual structural subsystems of a large-scale structure, while the latter, in contrast to other alternatives (e.g. the Extended Kalman Filter), offers a number of advantages in treating nonlinear systems, such as a derivative free calculation and a capacity for handling higher order nonlinearities. To this end, after formulating a detailed large-scale FE model for the vehicle frame substructure, a lumped model is adopted for the description of the nonlinear suspensions. Accordingly, a joint state and parameter estimation (JS&PE) problem is formulated on the basis of the lumped model. The proposed framework uses acceleration measurements from a limited number of sensors attached on the structure and a UKF observer for fusing these with a nonlinear FE substructure model, resulting into a JS&PE problem. The results indicate the validity of the proposed framework and motivate further implementation in large-scale structural systems with nonlinear components Show more
Book titleProceedings of the 2nd International Conference on Uncertainty Quantification in Computational Sciences and Engineering
PublisherNational Technical University of Athens (NTUA)
Organisational unit03890 - Chatzi, Eleni
MoreShow all metadata