查看更多>>摘要：Biomedical images contain a large volume of sensor measurements, which can reveal the descriptors of the disease under investigation. Computer-based analysis of such measurements helps detect the disease, and thereby swiftly aid medical professionals to choose adequate therapy. In this paper, we propose a robust deep learning ensemble framework known as COVID Fuzzy Ensemble Network, or COFE-Net. This strategy is proposed for the task of COVID-19 screening from chest X-rays (CXR) and CT Scans, as a part of Computer-Aided Detection (CADe) for medical practitioners. We leverage the strategy of Transfer Learning for Convolutional Neural Networks (CNNs) widely adopted in recent literature, and further propose an efficient ensemble network for their combination. The principles of fuzzy logic have been leveraged to combine the measured decision scores generated by three state-of-the-art CNNs - Inception V3, Inception ResNet V2 and DenseNet 201 - through the Choquet fuzzy integral. Experimental results support the efficacy of our approach over empirical ensembling, as the fuzzy ensembling strategy for biomedical measurement consists of dynamic refactoring of the classifier ensemble weights on the fly, based upon the confidence scores for coalitions of inputs. This is the chief advantage of our biomedical measurement strategy over others as other methods do not adjust to the multiple generated measurements dynamically unlike ours.Impressive results on multiple datasets demonstrate the effectiveness of the proposed method. The source code of our proposed method is made available at: https://github.com/theavicaster/covid-cade-ensemble.
de Medeiros, Luiz H. A.Coutinho, Marcelo S.Barbosa, Douglas C. P.Tarrago, Vinicius L....
10页
查看更多>>摘要：Metallic anchor rods are used to fix transmission tower guy wires. In order to perform visual inspections on the integrity of the anchor rods, electricity companies are obliged to undertake soil excavation, which is both costly and hazardous. Despite this, Frequency domain reflectometry (FDR) may be used to detect faults on anchor rods, and is enabled by a robust, high-frequency microwave device for support and connection. Impedance changes occur along the rod due to faults, thereby causing reflections, which are detected during FDR measurement analysis. However, auxiliary structures on the anchorage system also provide similar results during the analysis, and may therefore be confused with faults. An innovative, new approach is proposed in order to minimize the effects of small structures positioned at a distance from the fault, so that the fault reflection remains the same. Preliminary results have indicated reductions of up to 95% of the undesired reflection values.
查看更多>>摘要：Research on the optimization of the frequency bandwidth of calorimetric anemometers with hot-wire sensors, intended for the study of fast-changing and pulsed flows, is undergoing constant progress. The miniaturization of sensors reduces the thermal inertia of the measuring element, the structure of the electronic system and parameters of its elements are optimized, methods and algorithms for processing the output signal are developed. In a classic hot-wire anemometer, a single analog signal from the system is converted into a signal proportional to the measured velocity using the static calibration function of the anemometer. The article describes a method that simultaneously uses two measurement signals from an anemometer, proportional to the voltage and current of the sensor. Based on these signals, the measured flow velocity is calculated using a dynamic model of the sensor. This allows the measurement system's bandwidth to be extended and dynamic errors to be minimized.
查看更多>>摘要：Frequency response analysis is one of the best methods to diagnose mechanical defects in the power transformer winding. In this paper, sensitivity analysis of four components of the transfer function (TF), i.e., amplitude, phase, real and imaginary components of TF is investigated for different defects. The classification of type, location, and extent of different defects is carried out using support vector machine with cubic polynomial kernel. The accuracy of different indices and four different components of TF are investigated. Moreover, a new index based on fitting performance is defined. The results show that despite widespread use and popularity of amplitude of TF, the phase, imaginary, and real components of TF, have more sensitivity to the defects respectively and are more accurate for classification. Also, compared with the results of some high usage indices, the proposed index has the best classification accuracy.
查看更多>>摘要：The computer aided detection of diesel multiple properties is an active field of energy and chemical research as a result of the need for quality control and brands management of diesel raw materials. Based on this premise, this paper aimed to detect the diesel density, viscosity, freezing point, boiling point, cetane number and total aromatics using near infrared spectroscopy (NIRS) data combined with improved XY co-occurrence distance (ISPXY) and improved grey wolf optimized support vector regression (IGWO-SVR). The outcomes of average recovery, mean square error, mean absolute percentage error and determination coefficient of the proposed model are all better than other machine learning models. Further, this method is green, simple, effective, rapid, and can be embedded in the industrial network as a unit, which provides intelligent guidance for refineries to accurately control the quality of diesel oil.
Pacheco-Estrada, A. H.Hernandez-Marquez, F. L.Rodriguez-Resendiz, J.Duarte-Galvan, C....
7页
查看更多>>摘要：Scanners are tools for various measurement systems that eliminate human errors and automate measurement processes. This article reviews the scanners focused on low-frequency metrological measurements. The undesirable effects of the diverse types of scanners are described, and methods to characterize them and reduce their impact on measurement systems are specified. In the scope of the present paper an evaluation of the different types of scanners is carried out, highlighting their main characteristics. Additionally, a list of scanner design considerations for typical low-frequency electrical metrology applications is given.
查看更多>>摘要：Time-frequency (TF) features of nonstationary vibrations are indicative of the health condition of rotating machinery and, are also pivotal in analyzing acoustic signals obtained from processes such as bat echo-location. However, the TF features in these nonstationary vibration and acoustic signals are often submerged by strong background noise. This article proposes using the refined matching linear chirplet transform (RMLCT) to enhance the TF features, where the chirp rates are adaptively determined by spectral kurtosis and only the interesting time-frequency representations (TFRs) are retained. With selected chirp rates, a novel transformation kernel is developed, enabling the proposed method to simultaneously process nonstationary multicomponent signals. Moreover, the angle refinement strategy is proposed to improve the noise-handling capability of the proposed method. The signal reconstruction of the RMLCT is also analyzed, demonstrating that signal components of interest can be accurately reconstructed. Numerical and experimental analyses validate the effectiveness of the proposed RMLCT.
查看更多>>摘要：The transient electromagnetic method (TEM) is one of the most effective geophysical methods for detecting urban underground space. The underground resistivity distribution obtained by this imaging technique is important for engineers in urban construction and planning. The conventional transdimensional Bayesian (TransBayes) imaging method gives the credible interval of the data model and quantifies the uncertainty in the imaging results. However, due to the large number of calculations needed and the slow speed of TEM forward modelling, the engineering practicality of Trans-Bayes method is poor. To solve this problem, we propose a fast Trans-Bayes imaging approach based on the Bayesian information criterion (BIC) and the adaptive Metropolis (AM) algorithm. Our simulation and field experiments show that this BIC-AM approach can greatly improve Trans-Bayes imaging efficiency while ensuring imaging accuracy. This method may provide fast uncertainty estimates for urban TEM data.
查看更多>>摘要：In the context of the aeronautics industry, aluminum alloy structural parts are manufactured in several stages, from forming processes and heat treatments to final machining. Some process steps may generate residual stresses. Thus, material removal during machining releases these residual stresses, which induces part deformation. Such deformations can lead to geometric nonconformity of the machined part. It is therefore essential to control this phenomenon. Due to the variability in residual stress distribution in each raw part, the modeling approaches must to be coupled with experimental measurements. This article thus aims to define a reliable experimental technique for measuring in-plane deformation of large aeronautical parts during their machining. The backbone of the technique relies on Digital Image Correlation (DIC), which enables the contactless measurement of part deformation during machining. Moreover, DIC provides a full-field measurement and a direct evaluation of part deformations. This work discusses more specifically problems related to the use of DIC during machining, the latter corresponding to a particularly harsh environment. Indeed, optical systems undergo undesirable movement and metal chips hide areas of the observed part. These unwanted events corrupt the results. In order to control these problems and consistently apply DIC part deformation measurement during machining, specific methods are proposed in this paper. Finally, DIC measurements are performed during the same machining sequence of two parts. The excellent agreement of the two measurements confirms the reliability of the technique. Finally, measurements are discussed, emphasizing the contribution they provide to the machining community.
查看更多>>摘要：Phased array transducers sweep ultrasonic beams electronically without moving the probe which can be used in multipath ultrasonic flowmeters to reduce the number of transducers. The path arrangements optimization and numerical integration improvement are based on a model for the flow field. Therefore, the accuracy of multipath ultrasonic flowmeters is strongly influenced when the flow field deviates from the assumed velocity model. In this paper, a generalized velocity model is presented to reduce the sensitivity of flow measurement to deviation of the flow field. The performance of multipath ultrasonic phased array flowmeter is investigated using Gaussian quadrature method based on the generalized model under different flow fields. The results reveal that the generalized model is able to reduce the sensitivity of multipath phased array flowmeters to flow field. The minimum mean of relative errors is within 1.64% under symmetric and asymmetric flow fields.
NSTL
Elsevier