Search

The thin cloud removal (CR) technique has great practical value for the application of remote-sensing images. Existing deep-learning-based methods have attained remarkable achievements. However, most of them neglect the inherent feature correlations in deeper layers due to learning successively. In this letter, we propose a compact thin CR network based on the feedback (FB) mechanism, called CRFB-Net, which leverages the high-level features as FB information to modulate shallow representations. CRFB-Net employs the recurrent architecture to achieve such an FB scheme. Specifically, the restoration process does not terminate after obtaining an output. In this case, the output of intermediate iterations will flow into the next iteration as FB. For better utilization of FB, a multiscale feature fusion block (MFFB) is designed to refine the low-level representations from three scales. Furthermore, we introduce a curriculum learning (CL) strategy to train the CRFB-Net by gradually increasing the complexity of restoration, through which a sharper result is produced step by step. Extensive experiments demonstrate the superiority of our CRFB-Net, outperforming state-of-the-art (SOTA). [ABSTRACT FROM AUTHOR]

Broadband transient current measurement is a basic research method in many engineering fields. Optical integrated current sensor has the advantages of small size, anti electromagnetic interference and high precision, which has become a major trend of current monitoring. However, the existing optical current sensing methods have great challenges in broadband, wide amplitude and volume miniaturization. In this paper, a new high-performance magneto-optical energy coupling sensing method based on silicon photons and magneto-optical materials is studied, and an integrated optical waveguide current sensor based on a magneto-optical thin film and a double-ring assisted Mach–Zehnder interferometer is proposed. The energy coupling between the micro-ring resonator and the magneto-optical thin film is achieved through the magneto-optical nonreciprocity effect, and the current information is obtained by modulating the output light intensity with the Mach–Zehnder interferometer. The experimental results show that the sensor has a good linearity and a large dynamic measurement range. The measurement frequency range is from10 Hz to 13.56 MHz, and the sensitivity is 2.183 mV/A. This work seeks to promote the development of optical broadband current sensors on chips. [ABSTRACT FROM AUTHOR]

Partial discharge (PD) online monitoring is an essential part of transformer insulation condition based maintenance and prognostics and health management in industrial power system. In order to realize online localization of multiple PD sources conveniently, an approach that uses single-channel mixed acoustic emission (AE) signal is proposed in this article. First, the combined method based on the singular spectrum analysis and the independent component analysis is used to calculate the number of PD sources and realize the separation of mixed AE signals. Next, a difference selection algorithm is proposed to solve the problem that the inherent disorder of signal blind source separation makes the time difference of arrival technology difficult to localize the multiple PD sources directly. Then, the simulation signals and measured signals have verified that the signal separation effect of the proposed single-channel-based method is no less than that of the multichannel-based method. Finally, the proposed approach is applied to a transformer. The disordered separated signals were successfully grouped to realize multisource PD localization. [ABSTRACT FROM AUTHOR]

Harvesting energy from the human body has attracted substantial attention from industries and academics as it is a promising solution to address the battery life issue. In this article, we developed a bimorph piezoelectric bending beam-based energy harvester to scavenge energy from the human knee motion for powering body-worn electronics such as smartwatches and health monitors. During walking, the human knee will flex and extend to periodically deform piezoelectric macro fiber composites (MFC) bounded to the bending beam, thus generating electricity. The mathematical models of the whole system including the model of the large deflection of the bending beam, movement analysis of the end of the bending beam (slider-crank mechanism), the model of piezoelectric transducers, and measurement of the human knee angle are developed to comprehensively study the harvester. To validate the presented mathematical models, experimental testing on a human body when the subject walks on a treadmill at three different walking speeds is conducted here. The experimental results show good agreement with the simulation results, which indicates that the developed models can characterize the harvester. In addition, the performance of the harvester walking in different contexts is also evaluated. The average power can reach 12.79 mW for level-ground walking, 9.91 mW for stair descending, and 7.70 mW for stair ascending, when walking at a self-selected speed. [ABSTRACT FROM AUTHOR]

With the development of Internet of Things technology and artificial intelligence, the electric field sensor is required to be miniaturized, high performance, and low cost. At present, the commercial use of electric field sensing technology is facing several limitations, such as dynamic range, bandwidth, size, and signal-to-noise ratio. In this paper, a novel integrated waveguide wide-band electric field sensing solution is proposed. Based on the evanescent field and electro-optic effects, the high efficiency electro-optic energy coupling between the micro-ring resonator and the electro-optic polymer film is realized. Further, the information of the applied electric field is obtained by detecting the output light intensity of the Mach-Zehnder interferometer. The sensor can measure the electric field efficiently over a wide frequency band (50 Hz–13.5 MHz) and maintains good temperature stability. The sensor is small in size, simple in structure, low in cost, and easy to integrate. In addition, the sensor has the capability to realize the functions such as atmospheric electric field monitoring and lightning warning, providing a technical foundation for the construction of the Internet of Things. [ABSTRACT FROM AUTHOR]

An accurate discrimination on moisture status of oil-impregnated paper (OIP) bushings is crucial for the maintenance and replacement schedule of bushings. Based on frequency-domain spectroscopy (FDS) measurement and Dissado–Hill (DH) relaxation model, this article proposes a hybrid approach of hidden Markov model and gray wolf optimization (GWO-HMM) for MS estimation of bushings subjected to the ununiform moisture distribution and dynamic time-series modeling. First, simulation models of moisture diffusion and FDS of the OIP bushing were constructed using finite element modeling approach. Then, the GWO algorithm was employed to explore dielectric parameters influenced by moisture in the DH model. Then, the GWO-HMMs was adopted as a classification tool to discriminate the MS of the simulation and experimental data of the bushings. Classification results confirm that the average identification accuracies of the proposed method are 97.17% and 96.81% over these two datasets, which demonstrates the effectiveness of the proposed moisture estimate method for OIP bushings. [ABSTRACT FROM AUTHOR]