A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Automatic modal identification via eigensystem realization algorithm with improved stabilization diagram technique
Highlights Proposed an improved stabilization-diagram technique, via combined usage of Monte-Carlo sampling, FCM clustering and three-stage sifting manipulations. Applied the proposed method to ERA and SSI to form automated modal identification methods. Examined the performance of proposed method via comparison with other techniques through both numerical and field studies. The improved EAR outperforms most existing methods from the perspective of computational efficiency and identification accuracy.
Abstract Efficient and automatic identification of modal parameters becomes increasingly important for real-time structural health monitoring (SHM) of civil structures. As spurious modes usually exist as a key problem for most output-based identification methods, great efforts have been made to eliminate them typically via stabilization-diagram techniques. However, the quality of traditional stabilization diagrams depends on preset parameters whose values usually vary from one case to another, which makes the method to be less objective and low efficient. This article proposes an improved stabilization-diagram technique, through combined usage of Monte-Carlo sampling simulation, as well as fuzzy C-means (FCM) clustering and three-stage sifting manipulations. While the Monte-Carlo simulation aims to generate more robust stable-axis, the sifting and clustering manipulations can further remove outliers and discriminate true modal results. The improved stabilization-diagram technique is then applied to two mainstream modal identification methods, i.e., eigensystem realization algorithm (ERA) and stochastic subspace identification (SSI) under the context of both a simulation study on a dynamic system and a field research about a super-tall building. Results through comparison demonstrate that the improved stabilization-diagram technique can facilitate ERA and SSI to identify modal parameters automatically and effectively at a comparably good accuracy. However, ERA outperforms SSI evidently in terms of computational efficiency (upmost 15 times faster), which is attractive for real-time SHM. Parametric analysis has been also conducted to examine detailed performance of ERA aided by the proposed stabilization-diagram technique. Overall, the aforementioned method can be adopted to achieve a good balance between identification effectiveness and computational efficiency in an automatic working pattern, and has application prospect for real-time SHM of civil structures.
Automatic modal identification via eigensystem realization algorithm with improved stabilization diagram technique
Highlights Proposed an improved stabilization-diagram technique, via combined usage of Monte-Carlo sampling, FCM clustering and three-stage sifting manipulations. Applied the proposed method to ERA and SSI to form automated modal identification methods. Examined the performance of proposed method via comparison with other techniques through both numerical and field studies. The improved EAR outperforms most existing methods from the perspective of computational efficiency and identification accuracy.
Abstract Efficient and automatic identification of modal parameters becomes increasingly important for real-time structural health monitoring (SHM) of civil structures. As spurious modes usually exist as a key problem for most output-based identification methods, great efforts have been made to eliminate them typically via stabilization-diagram techniques. However, the quality of traditional stabilization diagrams depends on preset parameters whose values usually vary from one case to another, which makes the method to be less objective and low efficient. This article proposes an improved stabilization-diagram technique, through combined usage of Monte-Carlo sampling simulation, as well as fuzzy C-means (FCM) clustering and three-stage sifting manipulations. While the Monte-Carlo simulation aims to generate more robust stable-axis, the sifting and clustering manipulations can further remove outliers and discriminate true modal results. The improved stabilization-diagram technique is then applied to two mainstream modal identification methods, i.e., eigensystem realization algorithm (ERA) and stochastic subspace identification (SSI) under the context of both a simulation study on a dynamic system and a field research about a super-tall building. Results through comparison demonstrate that the improved stabilization-diagram technique can facilitate ERA and SSI to identify modal parameters automatically and effectively at a comparably good accuracy. However, ERA outperforms SSI evidently in terms of computational efficiency (upmost 15 times faster), which is attractive for real-time SHM. Parametric analysis has been also conducted to examine detailed performance of ERA aided by the proposed stabilization-diagram technique. Overall, the aforementioned method can be adopted to achieve a good balance between identification effectiveness and computational efficiency in an automatic working pattern, and has application prospect for real-time SHM of civil structures.
Automatic modal identification via eigensystem realization algorithm with improved stabilization diagram technique
Feng, Wen-Hai (author) / Wu, Chao-Yuan (author) / Fu, Ji-Yang (author) / Ng, Ching-Tai (author) / He, Yun-Cheng (author)
Engineering Structures ; 291
2023-06-06
Article (Journal)
Electronic Resource
English
Spurious mode distinguish by eigensystem realization algorithm with improved stabilization diagram
| British Library Online Contents | 2017
| British Library Online Contents | 2011
Mode identification by eigensystem realization algorithm through virtual frequency response function
| Wiley | 2019
An improved mode accuracy indicator for Eigensystem Realization Analysis (ERA) techniques
| Online Contents | 2012