1,720,962 research outputs found
A Tiny Machine Learning Approach to the Edge Localization of Acoustic Sources via Convolutional Neural Networks
No full textSource localization is a critical step in Acoustic Emission (AE)-based Structural Health Monitoring (SHM), since it allows to identify the point of a structure where most of the acoustic activity is growing due to both ageing (e.g., cracks, delamination, etc.) and sudden flaws. Recently, Artificial Intelligence (AI) algorithms have been proposed, which can overcome standard statistical methods especially when the signal-to-noise ratio is poor. In this work, the embodiment of tiny Convolutional Neural Network (CNN) models on a 32-bit microcontroller unit is presented for the task of Time of Arrival (ToA) estimation, which is the crucial parameter to be estimated for AE localization. Experimental results on real-field data prove that the embedded models can achieve satisfying accuracy for AE identification
Evaluating the Effect of Intrinsic Sensor Noise for Vibration Diagnostic in the Compressed Domain Using Convolutional Neural Networks
No full textMachine learning allows designing intelligent sensing networks capable to perform automatic inferences about the integrity of technical facilities. Compression techniques decrease significantly energy requirements of the sensing networks proving essential when sensing nodes are not supported by constant power sources. Existing schemes pass through the reconstruction of the original time series data before moving to the diagnosis phase. However, this passage can be avoided, i.e., inference can be performed directly in the compressed domain, by exploiting the specific information retained in the compressed patterns. This paper fulfills the goal above in the context of vibration-based structural health monitoring by proving, from an empirical perspective, that Convolutional Neural Networks (CNNs) can be used to predict the structural health status directly in the compressed domain when properly combined with adapted Compressed Sensing mechanisms. Importantly, the study analyses the effect of the intrinsic noise that affects digital accelerometer sensors. Results confirm that CNNs can mine information in the compressed domain even in presence of strong noise components, i.e., accuracy remains above 94% even for ultra-low-cost solutions featuring a signal-to-noise-ratio below 20 dB
Vibration Monitoring in the Compressed Domain With Energy-Efficient Sensor Networks
No full textStructural health monitoring (SHM) is crucial for the development of safe infrastructures. Onboard vibration diagnostics implemented by means of smart embedded sensors is a suitable approach to achieve accurate prediction supported by low-cost systems. Networks of sensors can be installed in isolated infrastructures allowing periodic monitoring even in the absence of stable power sources and connections. To fulfill this goal, the present letter proposes an effective solution based on intelligent extreme edge nodes that can sense and compress vibration data onboard, and extract from it a reduced set of statistical descriptors that serve as input features for a machine learning classifier, hosted by a central aggregating unit. Accordingly, only a small batch of meaningful scalars needs to be outsourced in place of long time series, hence paving the way to a considerable decrement in terms of transmission time and energy expenditure. The proposed approach has been validated using a real-world SHM dataset for the task of damage identification from vibration signals. Results demonstrate that the proposed sensing scheme combining data compression and feature estimation at the sensor level can attain classification scores always above 94%, with a sensor life cycle extension up to 350× and 1510× if compared with compression-only and processing-free implementations, respectively
Combining Compressed Sensing and Neural Architecture Search for Sensor-Near Vibration Diagnostics
No full textCompressed sensing (CS) for sensor-near vibration diagnostics represents a suitable approach for the design of network-efficient structural health monitoring systems. This article presents a solution for vibration analysis based on deep neural networks (DNNs) trained on compressed data. The envisioned maintenance system consists of a network of sensing nodes orchestrated by a very constrained centralizing unit. The latter is equipped with a microcontroller unit (MCU) that predicts the health state using the aggregated information. As a major contribution, the DNN architectures are generated automatically from the data through a procedure inspired by hardware-aware (HW) neural architecture search (NAS), called as HW-NAS-CS, which is uniquely refined with additional constraints that consider both the peculiarities of CS parameters and the limitation of embedded devices. The proposed approach has been validated using two real-world SHM datasets for vibration damage identification and eventually deployed on a low-end computing platform (the STM32L5 MCU). Results demonstrate that DNNs combined with adapted CS schemes can attain classification scores always above 90% even in case of very huge compression levels (higher than 64x): these performances significantly improve the ones attained by state-of-the-art approaches in the field, with the utmost advantage of being portable on embedded devices
Low depth time reversal modulation technique for ultrasonic guided waves-based communications
No full textUltrasonic guided waves (GWs) have the peculiar capability to travel long distances with minimal attenuation; thus, they represent a compelling alternative which exploits the mechanical waveguide as a form of communication channel and the elastic waves as the signals carrying information. The objective of this work is to implement an effective GWs-based communication system which takes advantage of the channel reciprocity to counteract reverberations and multi-path fading inherent to the dispersive nature of GWs, simultaneously compensating the in-operation mutual interferences between active transducers. More in detail, a strategy based on the Time-Reversal (TR) method is considered, which is combined in an original fashion with a low-depth synthesis of the time-reversed waveforms to be compatible with low-cost switching amplifiers. Experiments performed by simulating the propagation in Aluminum plates suggested that a communication rate up to tenth of kHz can be achieved in presence of highly digitized wave-forms without losing the original information content
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
A Damage Detection Strategy Based on Autoregressive Parameters
Full text linkStructural Health Monitoring (SHM) based on Operational
Modal Analysis (OMA) is pivotal in assessing the integrity of structures
and infrastructures in dynamic regimes. However, the successful extraction
of modal parameters and damage indexes through OMA typically
relies on a dense network of sensors working synchronously. This research
aims at alleviating this issue by resorting to autoregressive (AR) models
computed at individual sensing locations for damage detection, paving
the way to a fully decentralized monitoring approach. Such framework,
in which sensors can extract AR parameters in an independent manner,
is explored to alleviate the need for strict data synchronization,
which is instead a typical requirement of OMA procedures. The Mahalanobis
distance is then used in combination with the Receiver Operating
Curve (ROC) as a damage indicator to identify potential anomalies
upon aggregating the collected sets of AR features from different sensors.
The methodology has been applied to a numerical model and a real
steel bridge, comparing the performance of the proposed damage detection
strategy with a traditional approach based on modal parameters.
Results demonstrate that the proposed AR-based procedure can be very
competitive over a pure natural frequency-driven alternative, reaching a
classification score as high as 98% in both scenarios
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Full text linkWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
- …
