Recursive stochastic subspace identification for structural parameter estimation

C. C. Chang; Z. Li

doi:10.1117/12.815422

Translator Disclaimer

30 March 2009 Recursive stochastic subspace identification for structural parameter estimation

C. C. Chang, Z. Li

Author Affiliations +

Proceedings Volume 7292, Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2009; 729235 (2009) https://doi.org/10.1117/12.815422
Event: SPIE Smart Structures and Materials + Nondestructive Evaluation and Health Monitoring, 2009, San Diego, California, United States

Abstract

Identification of structural parameters under ambient condition is an important research topic for structural health monitoring and damage identification. This problem is especially challenging in practice as these structural parameters could vary with time under severe excitation. Among the techniques developed for this problem, the stochastic subspace identification (SSI) is a popular time-domain method. The SSI can perform parametric identification for systems with multiple outputs which cannot be easily done using other time-domain methods. The SSI uses the orthogonal-triangular decomposition (RQ) and the singular value decomposition (SVD) to process measured data, which makes the algorithm efficient and reliable. The SSI however processes data in one batch hence cannot be used in an on-line fashion. In this paper, a recursive SSI method is proposed for on-line tracking of time-varying modal parameters for a structure under ambient excitation. The Givens rotation technique, which can annihilate the designated matrix elements, is used to update the RQ decomposition. Instead of updating the SVD, the projection approximation subspace tracking technique which uses an unconstrained optimization technique to track the signal subspace is employed. The proposed technique is demonstrated on the Phase I ASCE benchmark structure. Results show that the technique can identify and track the time-varying modal properties of the building under ambient condition.

Citation Download Citation

C. C. Chang and Z. Li "Recursive stochastic subspace identification for structural parameter estimation", Proc. SPIE 7292, Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2009, 729235 (30 March 2009); https://doi.org/10.1117/12.815422

ACCESS THE FULL ARTICLE