1,720,963 research outputs found
A cointegration-based approach for automatic anomalies detection in large-scale structures
Full text linkIn recent years, the development of structural health monitoring (SHM) solutions for the automatic evaluation of the health state of engineering structures is continuously growing. However, when considering real-world applications, structures are highly influenced by meteorological variations or human activities (like temperature, wind and traffic loading) which can overwhelm the changes induced by a damage. Thanks to its ability to remove the long-term trends from a set of variables of the same process, cointegration, a technique born in the field of econometrics, has been introduced about ten years ago in SHM applications as a valid method to project out the confounding influences, such as environmental and operational variations. Because of the few examples of implementation currently available, this paper provides an in-depth review of all the relevant aspects to consider when cointegration is used as damage detection strategy and data are acquired from real-world structures of large dimensions. The methodology is applied for the first time on a complex structure of a singular nature, i.e. the steel roof of the G. Meazza stadium in Milan, which consists of multiple modular elements referred to as rafts. The time series which measures the rotations of the rafts are used as input data for the development of the cointegration-based method. Then, Johansen procedure is adopted to create a unique feature from the multivariate dataset, namely cointegrating residual, in which the effects of environmental and operational variables are suppressed, while the effects due to damage remain evident. The obtained residual is therefore used for novelty detection by means of a control chart, demonstrating its effectiveness into identifying the presence of anomalies or modifications in the structure in a clear and timely manner
Image deconvolution techniques for motion blur compensation in DIC measurements
No full textDigital image correlation (DIC) measurements are affected by several sources of uncertainty. Motion blur is one of the most relevant problems in dynamic DIC applications. This work deals with the problem of compensating motion blur effects on DIC. Firstly, a robust motion blur estimation technique based on cepstral analysis is presented and validated. Secondly, the problem of image restoration has been tackled. Two image deconvolution techniques are presented: one based on cepstrum deconvolution and the other based on Wiener filter. The latter has shown better robustness in presence of noise. Each presented technique has been tested with synthetic DIC experiments. Results demonstrate that both the compensation algorithms are able to improve the accuracy of DIC measurement in presence of motion blur
Assessment of a Vision-Based Technique to Estimate the Synchronization of Jumping Crowds in Civil Structures
No full textVibration serviceability criteria for civil structures follow a three-step framework, namely the excitation source, the path and the receiver. The first step, which is also the focus of this study, deals with the characterization of human-induced loads. However, the design models reported in the current guidance and codes are very often overly conservative and cannot adequately represent the real nature of crowd excitation. In this work, we present a computer vision technique, based on the use of Digital Image Correlation (DIC), as a solution to this problem. In addition to a cheap and an easy to install set up, the system can provide a comprehensive assessment of the coordinate motion induced by occupying crowds of various sizes. To demonstrate the efficacy of the proposed method, the measured DIC data are compared to those coming from the accelerometers installed on multiple subjects while performing jumping activities on a real grandstand. Then, the vision-based approach is used to study and to quantify the level of synchronization among the individuals for a range of songs and metronome beats. Results demonstrate that the DIC technique achieves similar performance as the inertial sensors but overcomes some practical limitations related to these traditional systems
Vision-Based Method to Measure the Synchronization Level of Jumping Crowds
Full text linkThe prediction of the dynamic loads produced by groups of people is a crucial aspect of the design of stadiums or entertainment venues. This is because the coordinated motion of lively crowds may induce severe vibration levels in the structure, which can become critical for both human comfort and structural safety. However, the available information on this topic is very limited. Human loads often rely on deterministic models that do not consider the interaction and the coordination achieved by the participants or try to account for them through empirical assumptions. Therefore, they could find very little correspondence in realistic scenarios. This article aims to close this gap by introducing a vision-based technique able to directly measure crowd loading and quantify the synchronization level between individuals. Starting from a sequence of images of a jumping crowd, digital image correlation (DIC) is used to extract the vertical velocity of different regions occupied by the participants; then, the vertical force time record is estimated. Finally, the comparison between the actual force signals and their envelopes allows for estimating the crowd synchronization over time. The method has been successfully validated with two field tests on the grandstands of the Giuseppe Meazza stadium in Milan, demonstrating its ability to reliably estimate the synchronization level reached by the participants
Classification reliability of 3D shapes using neural networks in case of partial and noisy models
No full textDeep learning is nowadays a mature technique and it is widely applied to image processing and classification. In recent years, many authors tried to extend this approach also to 3D object classification. However, most of the works in this field refers to complete models, while in many real applications the single acquisition with a vision system may only provide a partial object representation. Thus, the main goal of this work is to study the behaviour of classification neural networks when partial 3D models are considered. In particular, the analysis is focused on the classification reliability using partial point clouds, evaluating the influence of noise level and object scaling on the overall network performance. Tests are carried out both on synthetic point clouds, generated by simulation of common acquisition techniques, and on real clouds acquired by a Kinect device. This pushes towards the development of hybrid solutions, where training is made on simulated clouds and the testing takes place on real scanned objects, providing interesting suggestions for practical applications
Estimation and compensation of motion blur for the reduction of uncertainty in DIC measurements of flexible bodies
No full textDigital image correlation (DIC) is a useful technique to measure displacement/strain fields both for static and dynamic problems in experimental mechanics. When monitoring moving objects with digital cameras, motion blur may occur if the shutter time reaches the time scale of the motion of the measurand. Consequently, motion blur is one of the most relevant problems in those dynamic DIC applications where shutter time cannot be set arbitrarily. This work deals with the problem of compensating motion blur effects on a generic DIC image. The problem of motion blur compensation to reduce DIC uncertainty is discussed in literature in the case of rigid target, where the amount of motion blur is the same for the whole image area. Deformable targets, instead, pose the additional problem when motion blur is variable within a single frame. In this paper a subset-based technique is proposed to estimate and compensate the motion blur for each image region. The approach is tested on synthetically deformed and blurred images of a notched beam specimen
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Development, Validation and Preliminary Experiments of a Measuring Technique for Eggs Aging Estimation Based on Pulse Phase Thermography
Full text linkAssessment of the freshness of hen eggs destinated to human consumption is an extremely important goal for the modern food industry and sale chains, as eggs show a rapid natural aging which also depends on the storage conditions. Traditional techniques, such as candling and visual observation, have some practical limitations related to the subjective and qualitative nature of the analysis. The main objective of this paper is to propose a robust and automated approach, based on the use of pulsed phase thermography (PPT) and image processing, that can be used as an effective quality control tool to evaluate the freshness of eggs. As many studies show that the air chamber size is proportional to the egg freshness, the technique relies on the monitoring of the air chamber parameters to infer egg aging over time. The raw and phase infrared images are acquired and then post-processed by a dedicated algorithm which has been designed to automatically measure the size of the air chamber, in terms of normalized area and volume. The robustness of the method is firstly assessed through repeatability and reproducibility tests, which demonstrate that the uncertainty in the measure of the air chamber size never exceeds 5%. Then, an experimental campaign on a larger sample of 30 eggs, equally divided into three size categories (M, L, XL), is conducted. For each egg, the main sizes of the air chamber are measured with the proposed method and their evolution over time is investigated. Results have revealed, for all the egg categories, the existence of an analytic relationship and a high degree of correlation (R-2 > 0.95) between the geometric data of the air chamber and the weight loss, which is a well-known marker of egg aging
Damage Detection Using Supervised Machine Learning Algorithms for Real-World Engineering Structures
No full textVibration-based structural health monitoring represents an efficient way to evaluate structural integrity and the presence of damage at an early stage. These methods usually assume that damage manifests itself as a deviation in the modal properties of the structure with respect to its normal conditions. Traditional procedures for modal parameters estimation require the use of a dense sensor arrangement and complex logic techniques, thus making them not particularly suitable for the case of large engineering structures, where the need for cost-effective monitoring solutions is of utmost importance because of the large number of substructures to be monitored. This paper proposes the use of simple statistical and spectral features as a mean to characterize accelerations signals. Starting from this set of features, the principal component analysis (PCA) is first used to reduce data dimensionality still preserving the relevant information about the structural conditions, then a k-Nearest Neighbors (k-NN) procedure is adopted as a supervised machine learning method to classify different types of damage. The procedure is validated using the experimental data from the permanent monitoring system of the G. Meazza stadium grandstands of, where one accelerometer per stand is installed to get vibration data during the main events. Four grandstands located on the same ring and having the same nominal geometry are considered. The leading idea is to reproduce different scenarios where, due to the impossibility of imposing realistic damages, one grandstand is assumed to be the safe structure, while the others represent a proxy for small structural changes to be identified
- …
