查看更多>>摘要：? 2021In the laboratory, using traditional methods to measure the resonant frequency of transducers is inefficient and costly. For this, in this study, the feature variables determining the resonant frequency of transducer was firstly analyzed with the equivalent circuit method. Based on the input–output mapping function of back propagation (BP) neural network, the nonlinear relationship model between feature variables and the resonant frequency was established. Then, particle swarm optimization (PSO) was used to optimize the initial weight and threshold of BP neural network to obtain the best parameter combination, thereby improving the output accuracy of BP neural network. The experimental data showed that it is effective to measure the resonant frequency of the transducer by PSO-BP neural network, and the measurement error was 1 Hz. Compared to traditional measurement method, the intelligent method has the advantages of low cost, high efficiency, fast response speed, and strong robustness.
查看更多>>摘要：? 2022 Elsevier LtdTurbulence intensity levels are assigned different values according to set a probability distribution. Many researchers do not consider probability distributions when determining turbulence intensity levels. This is one cause of the disagreement between numerical simulations of turbulence modeling and experimental data. This study focuses on the use of weighted statistics to evaluate turbulence intensity levels through three types of examples. The first is a hands-on example with a Pitot-static tube. The second utilizes numerical data and the Monte Carlo method. The third relies on a dataset by constant temperature anemometry and laser Doppler anemometry, available from the literature. This study finds that turbulence intensity evaluations differ depending on the selection of the probability distribution. The turbulence intensity according to a Gaussian distribution is reduced by more than 30% in comparison with that set with a rectangular distribution. The notion of the true and engineering turbulence intensity is introduced.
查看更多>>摘要：? 2021High-pressure steel wire braided hoses have periodic braided steel wire structure and its productivity is too huge to refrain from making mistakes occasionally such as wire missed wire stacked and wire loosened. The major sticking points in detection are ultra-small, structural similarity and high reflectance of the steel wire braided hoses. Therefore, this paper proposed an automated method based on YOLOv5 to replace traditional naked eyes detection. The anchor boxes were optimized by K-means++ algorithm for defects dataset. Focal loss was used to mitigate the impact of samples imbalance. The predict heads were modified to increase the detection accuracy of ultra-small targets with guaranteeing speed. The effect of efficient channel attention mechanism (ECA) in different positions was analyzed to achieve a better network and increase detection performance. The experimental results showed that the accuracy and recognition efficiency of the proposed modified-YOLOv5s model could reach 92.2% and 23 frames per second.
查看更多>>摘要：? 2021This paper presents a 12-bit retransmission-based chipless Radio Frequency Identification (RFID) in the frequency range of 1.3–2.8 GHz. The designed tag is fully planar and is realized on a low-cost FR-4 substrate with a relative permittivity of 4.4 (loss tangent of 0.0018), and thickness of 1.6 mm. The overall footprint of the tag is 20 × 24 × 1.6 mm3, and it contains two high flat gain orthogonal antennas for transmitting and receiving a long-read range of data. The structure consists of a modified transmission line to form a square loop, strips of different lengths arranged in a readable configuration, and strips to enable tunable frequency operation. The flower-shaped antennas use a partial ground with a modified circular disc-shaped patch to attain desirable bandwidth and maximum flat gain. The frequency coding technique increases the coding density using three resonators, with 10 combinations providing 1000 (103) identification codes. The measured results of the designed data-encoding circuit correlate with the simulated results, ensuring the reliability of real-time low-profile applications.
查看更多>>摘要：? 2022 Elsevier LtdThe present paper aims to describe the project and development of an ECG monitoring system which is able to diagnose specific cardiac pathologies by adapting the processing algorithm to the monitored patient. The system can work standalone by providing the final diagnosis to the patient by a LEDs set. In addition, an IoT based architecture allows the system to share data and diagnosis with a remote cardiologist in real-time or to store data in a FTP folder, as an Holter monitor, for post-processing and further analysis. The system consists of two sections: the digital one, based on the National Instruments MyRIO, digitalises the signal and processes it in real-time to make a diagnosis on possible occurring cardiac pathologies (bradycardia, tachycardia, infarction, ischemia); the analog section, which performs the acquisition, amplification, and preliminary filtering of a 3-leads electrocardiographic signal. The processing algorithm has been developed in NI LabVIEW environment. The main contribution to the state of the art is due to two removable and updatable memory devices. A first memory has been used to store clinical and personal data of patient so to configure the computing algorithm and adapt it to the monitored patient. A second memory has been used to store accuracy and uncertainty specifications of the ECG system in order to improve the reliability of the final diagnosis. The system aims to provide an electrocardiographic monitoring system for healthcare applications in the smart city context. The possibility to monitor constantly the patient health state at home is an important challenge of the next future smart cities. IoT and automation are the main aspects of the proposed system. Patients having chronic heart diseases need frequent hospitalizations to check their heart health. In this scenario, the proposed system is an Ambient Assisted Living solution developed to encourage the independent life of cardiac subjects by supporting his/her care and wellness during the daily life.
查看更多>>摘要：? 2022 Elsevier LtdThe distribution of the power transformers at a far distance from the electrical plants represents the main challenge against the diagnosis of the transformer status. This paper introduces a new integration of an Internet of Things (IoT) architecture with deep learning against cyberattacks for online monitoring of the power transformer status. A developed one dimension convolutional neural network (1D-CNN), which is characterized by robustness against uncertainties, is introduced for fault diagnosis of power transformers and cyberattacks. Further, experimental scenarios are performed to confirm the effectiveness of the proposed IoT architecture. While compared to previous approaches in the literature, the accuracy of the new deep 1D-CNN is greater with 94.36 percent in the usual scenario, 92.58 percent when considering cyberattacks, and ±5% uncertainty. The proposed integration between the IoT platform and the 1D-CNN can detect the cyberattacks properly and provide secure online monitoring for the transformer status via the internet network.
查看更多>>摘要：? 2022 Elsevier LtdTo improve the measurement accuracy of the double flank worm gear pair rolling tester, the grating system is calibrated by laser interferometer and then compensation model using polynomial function is established, uncertainty budget is carried out based on the sensitivity test and error sources analysis, and some factors are adjusted to lower the measurement uncertainty. After compensation, the positioning error of worm gear moving slider, worm slider and worm gear measuring slider is reduced from 1.80 μm to 0.20 μm, 20.00 μm to 0.50 μm, and from 0.70 μm to 0.05 μm respectively. Related sensitivity coefficients for measurement results total radial composite deviation Fid, tooth-to-tooth radial composite deviation fid,andrunoutFr are within 0.01 to 0.04. According to GUM, three parts are considered in uncertainty evaluation including standard uncertainty of base unit, standard uncertainty of data acquisition & software system and standard uncertainty of measurement condition. After the optimization of the main error sources, the expanded uncertainty for measurement results Fid, fid,Fr is reduced from 5.4 μm, 4.6 μm and 2.5 μm to 2.6 μm, 2.1 μm and 1.5 μm respectively. The experimental results show that, the accuracy improvement method for double flank worm gear pair rolling tester can reduce measurement uncertainty, and the whole work can be a certain reference for the accuracy improvement of precision instruments.
查看更多>>摘要：? 2022 Elsevier LtdThe locally optimized chirplet spectrogram (LOCS) is a novel method proposed in this work for generating a high-resolution and cost-effective spectrogram of the induction machine (IM) current, suitable for the identification of fault-related harmonics in transient conditions. Its distinctive novelty is that it optimizes automatically the parameters of the analysing window used for building the current spectrogram, at each point of the time–frequency plane, with the computational cost of a conventional, non-optimized spectrogram. It is based on the definition of a large dictionary of different chirplet windows, which are combined into a single, complex time window. A single short time Fourier transform with this new window generates in parallel the spectrograms of all the dictionary windows, and the LOCS chooses among them the locally optimized values in an automatic way. The proposed technique is applied to the diagnosis of two commercial induction motors with bar breakages and mixed eccentricity faults.
查看更多>>摘要：? 202177 GHz radar has become a promising approach to enhance automotive safety by quickly detecting and identifying targets around the vehicle, especially in harsh weather conditions. This requires 77 GHz radars to provide environmental imaging with high resolution and reliability. However, radar images are easily blurred by sidelobes and background noises, which makes it difficult to extract real target information. In this paper, a sidelobe suppression algorithm based on the point spread function (PSF) and complex-valued neural network has been proposed to discriminate and suppress unwanted sidelobes while maintaining mainlobes referring to targets. To overcome the scarcity of real-world 77 GHz multiple-input multiple-output (MIMO) radar datasets, this paper derives the formula of PSF for 77 GHz MIMO radars in detail and exploits the PSF to generate simulated datasets for training. In addition, to be compatible with the complex-valued radar datasets, a customized neural network model has also been established in this paper. The well-trained neural network is further adopted to suppress sidelobes on real-world radar images. Comprehensive simulations and measurements have proved the superior performance of the proposed method, especially in cases with low signal-to-noise ratio (SNR), large channel mismatches, small target Radar Cross-Section (RCS) and large observation angle.
查看更多>>摘要：? 2022 Elsevier LtdPerception over time is a critical problem in those cases where deformable objects are manipulated. The goal of this study is to cover the contour of an object along a deformation process and according to a prescribed coverage objective, in terms of visibility and resolution. This task is carried out by a set of limited field-of-view cameras. We propose novel methods for guaranteeing feasibility of the coverage objectives, which include the computation of the maximum visibility and resolution of the contour. Then, we introduce the coverage objectives in an offline constrained optimization problem to compute a priori the minimum number of cameras that achieve the coverage requirements. Finally, we propose an online technique that provides optimized configurations faster than the offline one, even when the object's reference deformation is unknown. We report experimental results in which our method achieves 100% of the coverage in simulation and in a real task.
NSTL
Elsevier