• Smart Structures and Systems
• /
• 제29권1호
• /
• pp.17-28
• /
• 2022

The cable component of cable-stayed bridges is gradually impacted by weather conditions, vehicle loads, and material corrosion. The stayed cable is a critical load-carrying part that closely affects the operational stability of a cable-stayed bridge. Damaged cables might lead to the bridge collapse due to their tension capacity reduction. Thus, it is necessary to develop structural health monitoring (SHM) techniques that accurately identify damaged cables. In this work, a combinational identification method of three efficient techniques, including statistical analysis, clustering, and neural network models, is proposed to detect the damaged cable in a cable-stayed bridge. The measured dataset from the bridge was initially preprocessed to remove the outlier channels. Then, the theory and application of each technique for damage detection were introduced. In general, the statistical approach extracts the parameters representing the damage within time series, and the clustering approach identifies the outliers from the data signals as damaged members, while the deep learning approach uses the nonlinear data dependencies in SHM for the training model. The performance of these approaches in classifying the damaged cable was assessed, and the combinational identification method was obtained using the voting ensemble. Finally, the combination method was compared with an existing outlier detection algorithm, support vector machines (SVM). The results demonstrate that the proposed method is robust and provides higher accuracy for the damaged cable detection in the cable-stayed bridge.

• Smart Structures and Systems
• /
• 제31권4호
• /
• pp.335-349
• /
• 2023

Bridges constantly undergo deterioration and damage, the most common ones being concrete damage and exposed rebar. Periodic inspection of bridges to identify damages can aid in their quick remediation. Likewise, identifying components can provide context for damage assessment and help gauge a bridge's state of interaction with its surroundings. Current inspection techniques rely on manual site visits, which can be time-consuming and costly. More recently, robotic inspection assisted by autonomous data analytics based on Computer Vision (CV) and Artificial Intelligence (AI) has been viewed as a suitable alternative to manual inspection because of its efficiency and accuracy. To aid research in this avenue, this study performs a comparative assessment of different architectures, loss functions, and ensembling strategies for the autonomous segmentation of bridge components and damages. The experiments lead to several interesting discoveries. Nested Reg-UNet architecture is found to outperform five other state-of-the-art architectures in both damage and component segmentation tasks. The architecture is built by combining a Nested UNet style dense configuration with a pretrained RegNet encoder. In terms of the mean Intersection over Union (mIoU) metric, the Nested Reg-UNet architecture provides an improvement of 2.86% on the damage segmentation task and 1.66% on the component segmentation task compared to the state-of-the-art UNet architecture. Furthermore, it is demonstrated that incorporating the Lovasz-Softmax loss function to counter class imbalance can boost performance by 3.44% in the component segmentation task over the most employed alternative, weighted Cross Entropy (wCE). Finally, weighted softmax ensembling is found to be quite effective when used synchronously with the Nested Reg-UNet architecture by providing mIoU improvement of 0.74% in the component segmentation task and 1.14% in the damage segmentation task over a single-architecture baseline. Overall, the best mIoU of 92.50% for the component segmentation task and 84.19% for the damage segmentation task validate the feasibility of these techniques for autonomous bridge component and damage segmentation using RGB images.

• Smart Structures and Systems
• /
• 제20권5호
• /
• pp.549-562
• /
• 2017

The US railroad network carries 40% of the nation's total freight. Railroad bridges are the most critical part of the network infrastructure and, therefore, must be properly maintained for the operational safety. Railroad managers inspect bridges by measuring displacements under train crossing events to assess their structural condition and prioritize bridge management and safety decisions accordingly. The displacement of a railroad bridge under train crossings is one parameter of interest to railroad bridge owners, as it quantifies a bridge's ability to perform safely and addresses its serviceability. Railroad bridges with poor track conditions will have amplified displacements under heavy loads due to impacts between the wheels and rail joints. Under these circumstances, vehicle-track-bridge interactions could cause excessive bridge displacements, and hence, unsafe train crossings. If displacements during train crossings could be measured objectively, owners could repair or replace less safe bridges first. However, data on bridge displacements is difficult to collect in the field as a fixed point of reference is required for measurement. Accelerations can be used to estimate dynamic displacements, but to date, the pseudo-static displacements cannot be measured using reference-free sensors. This study proposes a method to estimate total transverse displacements of a railroad bridge under live train loads using acceleration and tilt data at the top of the exterior pile bent of a standard timber trestle, where train derailment due to excessive lateral movement is the main concern. Researchers used real bridge transverse displacement data under train traffic from varying bridge serviceability levels. This study explores the design of a new bridge deck-pier experimental model that simulates the vibrations of railroad bridges under traffic using a shake table for the input of train crossing data collected from the field into a laboratory model of a standard timber railroad pile bent. Reference-free sensors measured both the inclination angle and accelerations of the pile cap. Various readings are used to estimate the total displacements of the bridge using data filtering. The estimated displacements are then compared to the true responses of the model measured with displacement sensors. An average peak error of 10% and a root mean square error average of 5% resulted, concluding that this method can cost-effectively measure the total displacement of railroad bridges without a fixed reference.

• Smart Structures and Systems
• /
• 제31권3호
• /
• pp.247-257
• /
• 2023

In structural health monitoring of large-scale structures, optimal sensor placement plays an important role because of the high cost of sensors and their supporting instruments, as well as the burden of data transmission and storage. In this study, a vibration sensor placement algorithm based on deep reinforcement learning (DRL) is proposed, which can effectively solve non-convex, high-dimensional, and discrete combinatorial sensor placement optimization problems. An objective function is constructed to estimate the quality of a specific vibration sensor placement scheme according to the modal assurance criterion (MAC). Using this objective function, a DRL-based algorithm is presented to determine the optimal vibration sensor placement scheme. Subsequently, we transform the sensor optimal placement process into a Markov decision process and employ a DRL-based optimization algorithm to maximize the objective function for optimal sensor placement. To illustrate the applicability of the proposed method, two examples are presented: a 10-story braced frame and a sea-crossing bridge model. A comparison study is also performed with a genetic algorithm and particle swarm algorithm. The proposed DRL-based algorithm can effectively solve the discrete combinatorial optimization problem for vibration sensor placements and can produce superior performance compared with the other two existing methods.

• Smart Structures and Systems
• /
• 제24권6호
• /
• pp.733-744
• /
• 2019

Wind speed forecasting is critical for a variety of engineering tasks, such as wind energy harvesting, scheduling of a wind power system, and dynamic control of structures (e.g., wind turbine, bridge, and building). Wind speed, which has characteristics of random, nonlinear and uncertainty, is difficult to forecast. Nowadays, machine learning approaches (generalized regression neural network (GRNN), back propagation neural network (BPNN), and extreme learning machine (ELM)) are widely used for wind speed forecasting. In this study, two schemes are proposed to improve the forecasting performance of machine learning approaches. One is that optimization algorithms, i.e., cross validation (CV), genetic algorithm (GA), and particle swarm optimization (PSO), are used to automatically find the optimal model parameters. The other is that the combination of different machine learning methods is proposed by finite mixture (FM) method. Specifically, CV-GRNN, GA-BPNN, PSO-ELM belong to optimization algorithm-assisted machine learning approaches, and FM is a hybrid machine learning approach consisting of GRNN, BPNN, and ELM. The effectiveness of these machine learning methods in wind speed forecasting are fully investigated by one-year field monitoring data, and their performance is comprehensively compared.

• Smart Structures and Systems
• /
• 제12권3_4호
• /
• pp.399-414
• /
• 2013

The smart sensor technology has opened new horizons for assessing and monitoring structural health of civil infrastructure. Smart sensor's unique features such as onboard computation, wireless communication, and cost effectiveness can enable a dense network of sensors that is essential for accurate assessment of structural health in large-scale civil structures. While most research efforts to date have been focused on realizing wireless smart sensor networks (WSSN) on bridge structures, relatively less attention is paid to applying this technology to buildings. This paper presents a decentralized damage detection using the WSSN for building structures. An existing flexibility-based damage detection method is extended to be used in the decentralized computing environment offered by the WSSN and implemented on MEMSIC's Imote2 smart sensor platform. Numerical simulation and laboratory experiment are conducted to validate the WSSN for decentralized damage detection of building structures.

• Smart Structures and Systems
• /
• 제10권2호
• /
• pp.131-154
• /
• 2012

In this paper, it is aimed to determine the seismic behaviour of highway bridges by nondestructive testing using ambient vibration measurements. Eynel Highway Bridge which has arch type structural system with a total length of 216 m and located in the Ayvaclk county of Samsun, Turkey is selected as an application. The bridge connects the villages which are separated with Suat U$\breve{g}$urlu Dam Lake. A three dimensional finite element model is first established for a highway bridge using project drawings and an analytical modal analysis is then performed to generate natural frequencies and mode shapes in the three orthogonal directions. The ambient vibration measurements are carried out on the bridge deck under natural excitation such as traffic, human walking and wind loads using Operational Modal Analysis. Sensitive seismic accelerometers are used to collect signals obtained from the experimental tests. To obtain experimental dynamic characteristics, two output-only system identification techniques are employed namely, Enhanced Frequency Domain Decomposition technique in the frequency domain and Stochastic Subspace Identification technique in time domain. Analytical and experimental dynamic characteristic are compared with each other and finite element model of the bridge is updated by changing of boundary conditions to reduce the differences between the results. It is demonstrated that the ambient vibration measurements are enough to identify the most significant modes of highway bridges. After finite element model updating, maximum differences between the natural frequencies are reduced averagely from 23% to 3%. The updated finite element model reflects the dynamic characteristics of the bridge better, and it can be used to predict the dynamic response under complex external forces. It is also helpful for further damage identification and health condition monitoring. Analytical model of the bridge before and after model updating is analyzed using 1992 Erzincan earthquake record to determine the seismic behaviour. It can be seen from the analysis results that displacements increase by the height of bridge columns and along to middle point of the deck and main arches. Bending moments have an increasing trend along to first and last 50 m and have a decreasing trend long to the middle of the main arches.

• Smart Structures and Systems
• /
• 제6권5_6호
• /
• pp.481-504
• /
• 2010

Wireless smart sensor networks (WSSNs) have been proposed by a number of researchers to evaluate the current condition of civil infrastructure, offering improved understanding of dynamic response through dense instrumentation. As focus moves from laboratory testing to full-scale implementation, the need for multi-hop communication to address issues associated with the large size of civil infrastructure and their limited radio power has become apparent. Multi-hop communication protocols allow sensors to cooperate to reliably deliver data between nodes outside of direct communication range. However, application specific requirements, such as high sampling rates, vast amounts of data to be collected, precise internodal synchronization, and reliable communication, are quite challenging to achieve with generic multi-hop communication protocols. This paper proposes two complementary reliable multi-hop communication solutions for monitoring of civil infrastructure. In the first approach, termed herein General Purpose Multi-hop (GPMH), the wide variety of communication patterns involved in structural health monitoring, particularly in decentralized implementations, are acknowledged to develop a flexible and adaptable any-to-any communication protocol. In the second approach, termed herein Single-Sink Multi-hop (SSMH), an efficient many-to-one protocol utilizing all available RF channels is designed to minimize the time required to collect the large amounts of data generated by dense arrays of sensor nodes. Both protocols adopt the Ad-hoc On-demand Distance Vector (AODV) routing protocol, which provides any-to-any routing and multi-cast capability, and supports a broad range of communication patterns. The proposed implementations refine the routing metric by considering the stability of links, exclude functionality unnecessary in mostly-static WSSNs, and integrate a reliable communication layer with the AODV protocol. These customizations have resulted in robust realizations of multi-hop reliable communication that meet the demands of structural health monitoring.

• 한국지진공학회논문집
• /
• 제10권6호
• /
• pp.115-123
• /
• 2006

본 논문에서는 구조건전성모니터링(SHM)시스템이 설치된 케이블교량을 대상으로 장기적인 동적거동 특성을 분석하였다. 3차원 유한요소모델을 사용한 모드해석을 통해 모드변수를 추출하였다. 이를 교량에서 측정된 상시진동신호에 대해 주파수영역에서 분석한 고유진동수와 비교하여 정확한 기저모델이 구축되었음을 알 수 있었다. 지난 5년간의 고유진동수와 온도를 통계분석하여 고유진동수가 온도에 선형 반비례하고 있음을 확인하였고, 이러한 온도효과에 대한 평가를 수행하였다. 또한 상시진동신호를 시간영역에서 TDD기법을 적용하여 모드형상을 추출하였으며, 모드해석 결과와의 비교를 통해 케이블교량에 적용이 가능함을 검증하였다.

• 대한토목학회논문집
• /
• 제31권4A호
• /
• pp.287-293
• /
• 2011

본 연구에서는 교량의 진동을 이용한 압전 외팔보 에너지 수확장치의 적용성을 연구하였다. 이를 위하여 압전 소자의 구성방정식과 외팔보의 진동방정식을 결합하여 외팔보의 단일 모드에 대한 연성 방정식을 행렬 형태로 구성하였다. 그리고 에너지 수확장치의 가진기 실험을 통하여 해석 모델의 타당성을 검증하였다. KTX, 새마을, 무궁화 열차가 주행할 때 측정된 교량 가속도를 바탕으로 수치해석을 통하여 산정한 에너지 수확장치의 최대 전력은 각각 28.5 mW, 0.65 mW, 0.51 mW로 나타났다. 이를 볼 때 철도와 같은 이동하중에 의한 교량의 진동은 가진 진동수와 가속도가 낮고 지속시간이 짧아서 에너지 공급원으로서 효율성이 떨어지는 것으로 판단된다.