“…This article extends preliminary results already published in Reference 44, where only the sensor selection algorithm and its application with the PCA on the bookshelf structure were proposed. In addition, in this article we extend the modeling with nonlinear components, that is, by including RTs and PE model structures, and validate the proposed methodologies on two additional experimental benchmarks.…”
In this article the problem of data‐driven structural damage detection is considered exploiting historical data collected from a structure. First, a novel technique based on Kalman filtering and on a combination of regression trees theory from machine learning and auto‐regressive system identification from control theory is derived to build switching models that can be used to detect structural damages. A technique is also proposed leveraging principal component analysis together with the poly‐exponential approach to create nonlinear models to be used for structural damage detection. Finally, a novel sensors selection algorithm based on the notions of entropy and information gain from information theory is developed to reduce the number of sensors without affecting or even improving, as it happens in our experimental setup, the model accuracy. The presented techniques are validated on three independent experimental datasets, showing that the proposed algorithms outperform previous and classical approaches, improving the prediction accuracy and the damage detection sensitivity while reducing the number of sensors.
“…This article extends preliminary results already published in Reference 44, where only the sensor selection algorithm and its application with the PCA on the bookshelf structure were proposed. In addition, in this article we extend the modeling with nonlinear components, that is, by including RTs and PE model structures, and validate the proposed methodologies on two additional experimental benchmarks.…”
In this article the problem of data‐driven structural damage detection is considered exploiting historical data collected from a structure. First, a novel technique based on Kalman filtering and on a combination of regression trees theory from machine learning and auto‐regressive system identification from control theory is derived to build switching models that can be used to detect structural damages. A technique is also proposed leveraging principal component analysis together with the poly‐exponential approach to create nonlinear models to be used for structural damage detection. Finally, a novel sensors selection algorithm based on the notions of entropy and information gain from information theory is developed to reduce the number of sensors without affecting or even improving, as it happens in our experimental setup, the model accuracy. The presented techniques are validated on three independent experimental datasets, showing that the proposed algorithms outperform previous and classical approaches, improving the prediction accuracy and the damage detection sensitivity while reducing the number of sensors.
“…Penyusunan sebuah ide untuk memilih sub himpunan data berdasarkan pada korelasi jarak, yang kemudian dapat digunakan sebagai referensi dalam posisi pemasangan sensor atau akselerometer di dalam sebuah struktur. Artikel ini disusun dengan formulasi sebagai berikut: Bab II akan membahas mengenai penelitian terkait dan dasar-dasar teori yang dibutuhkan, meliputi: metode KTP, korelasi jarak dan pemilihan sub himpunan berbasis entropi yang telah diajukan pada [13] dan [14]. Bab III akan membahas alur dan persamaan matematis dari algoritma yang diajukan pada penelitian ini dan data yang digunakan untuk memvalidasi performa dari algoritma tersebut.…”
Sering kali, sebuah kerusakan struktur yang masif terjadi karena pengabaian terhadap kerusakan kecil. Kejadian malang ini kemudian menimbulkan berbagai kerugiaan, baik secara material maupun korban jiwa. Oleh karena itu, dirasa penting untuk dapat mendeteksi kerusakan dari sebuah struktur sedini mungkin untuk mencegah terjadinya hal yang tidak diinginkan. Penelitian ini menggagas sebuah algoritma pendektesi kerusakan struktur bangunan berbasiskan pada metode korelasi jarak dan kuadrat terkecil parsial. algoritma ini berfokuskan pada pemilihan sekelompok sensor yang dapat bekerja secara optimal berdasarkan pada perhitungan korelasi jarak. Berdasarkan pada percobaan pada data experimental dari sebuah struktur jembatan, algoritma yang digagas dapat mengurangi jumlah akselerometer yang diperlukan hingga 80% untuk menyusun model prediktif tanpa mengurangi atau bahkan meningkatkan akurasi dari model prediktif akselerometer sebesar 1 hingga 1,3%. Lebih lanjut, algoritma yang digagas dapat mendekteksi keberadaan kerusakan struktur dengan baik, serta mampu mengkarakterisasi tingkat kerusakan dari struktur berdasarkan pada perubahan standar deviasi dari residu kuadrat.
“…Another instance of the application of entropy in sensor selection can be found in the field of structural damage detection. 23 The authors employed entropy as a metric to reduce the number of required sensors while maintaining the accuracy of the damage detection model. These examples highlight the potential of entropy as a valuable metric in the selection of optimal sensors.…”
Accurate and precise estimation of process variables is key to effective
process monitoring. The estimation accuracy depends on the choice of the
sensor network. Therefore, this paper aims at developing convex
optimization formulations for designing the optimal sensor network using
information-theoretic measures in linear steady-state data
reconciliation. To this end, the estimation errors are characterized by
a multivariate Gaussian distribution, and thus the analytical form for
entropy and Kullback-Leibler divergences (forward, reverse, and
symmetric) of estimation errors can be obtained to formulate the optimal
sensor network design. The proposed information theoretic-based optimal
sensor selection problems are shown to be integer semidefinite
programming problems where the relaxation of binary decision variables
results in solving a convex optimization problem. Thus, we use a branch
and bound method to obtain a globally optimal sensor network design.
Demonstrative case studies are presented to illustrate the efficacy of
the proposed optimal sensor selection formulations.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.