Search

In today's society, there is an urgent need to help users better access information that they are interested in, as there is an increasing amount of news and messages available with the development of the Internet. Many existing methods involve directly inputting text into a pre-trained model, which limits the effectiveness of text feature extraction. The personalized news recommendation model discussed in this article is a model that can enhance feature extraction from news articles. It consists of a candidate news module, a historically accessed news module, and an access prediction module. Using news titles that accurately summarize news content, a model with double multi-head attention mechanisms and double residual structures (DDM) is utilized to better capture the features of news articles historically accessed by users, thereby achieving an improved recommendation functionality. The candidate news module aims to help the model learn representations of news that users are likely to select from the news titles. The user historical click news module primarily serves to enable the model to learn personalized representations of users from news they have previously browsed. The model has been tested on MIND-small. The AUC reached 0.6665, the MRR reached 0.3205, the nDCG@5 reached 0.3532, and the nDCG@10 reached 0.4158. The results indicate this model has achieved good results in the downstream tasks of preprocessing news-title texts. [ABSTRACT FROM AUTHOR]

In this paper, a high-power 170 GHz frequency doubler based on a Schottky diode is proposed using an in-phase power-combining structure. Unlike a conventional power-combining frequency doubler, the proposed frequency doubler utilizes the combination of a T-junction power divider and two bend waveguides to eliminate the phase difference between the two output ports of the T-junction power divider, so as to achieve in-phase power combining with a concise structure. The frequency doubler was fabricated on a 50 μm thick AlN high-thermal-conductivity substrate to reduce the impact of the thermal effect on the performance. The measured results show that the doubler exhibits a conversion efficiency of 11–31.3% in the 165–180 GHz band under 350–400 mW of input power, and a 118 mW peak output power with a 31.3% efficiency was measured at 174 GHz 'when the input power was 376 mW. A good agreement was achieved between the simulation results and the measured performance of the doubler, which proves the effectiveness of the proposed in-phase power-combining structure. [ABSTRACT FROM AUTHOR]

HY-2C is the third satellite in China's ocean dynamic environment satellite series, and carries a correction microwave radiometer (CMR) to correct the wet tropospheric path delay for the aligned radar altimeter. To effectively use the brightness temperatures (TB) of CMR to retrieve path delay, an on-orbit calibration effort is required. In this study, an antenna pattern correction (APC) method and a neural network method are used to perform an on-orbit calibration for CMR's antenna temperatures and a model based on the Whale Optimization Algorithm (WOA), Levenberg–Marquardt (LM) algorithm, and Back-Propagation neural network (WOA–LM–BP) has been proposed to retrieve the wet tropospheric correction (WTC) of CMR. The on-orbit calibration results, compared with the simulated brightness temperatures calculated by the radiative transfer model (RTM), have shown that compared with the APC method, the neural network method can almost eliminate the latitude variation, and the total bias and standard deviation of the on-orbit calibrated TB at all channels have obviously decreased. The retrieved WTC results also have shown that the retrieved WTC of CMR has a good agreement with the corresponding ones from the model-derived WTC and Jason-3. [ABSTRACT FROM AUTHOR]

The voltage and current measurement of the medium-voltage (MV) side are used for the traditional fault location method, which leads to a high installation cost of the measurement and complicated post-operational and maintenance work. A fault location method is proposed based on the distributed measurement at the low-voltage (LV) side. On the analysis of voltage distribution rules and the influence of the distribution transformer on voltage transmission, obvious differences are found in the variation of voltage (phase voltage and sequence voltage) at the LV side for different faults—specifically, the detection sensitivity of the voltage to different faults varies. Therefore, a fault location method for the MV distribution network based on characteristic voltage at the LV side is proposed. Firstly, the characteristic voltage is selected adaptively according to the fault type. The suspected fault path is then determined by the characteristic voltage amplitude of measuring points. Finally, the fault segment is located via the characteristic current of each segment. This method can be applied in combination with the widely used LV measurement equipment such as the power consumption information acquisition system and the intelligent fusion terminal to acquire LV side voltage without adding new measurement devices. The distribution difference of the characteristic voltage at the LV side is applied for accurate fault finding, which is an economical and practical new idea for the fault location of the distribution network. The correctness and effectiveness of the method were verified by the simulation of the IEEE 34 system built in PSCAD/EMTDC and a real distribution network. [ABSTRACT FROM AUTHOR]

Terahertz scattering-scanning near-field optical microscopy (THz s-SNOM), combining the best features of terahertz technology and s-SNOM technology, has shown unique advantages in various applications. Consequently, building a model to characterize near-field interactions and investigate practical issues has become a popular topic in THz s-SNOM research. In this study, a finite element model (FEM) is proposed to quantify the near-field interactions, and to investigate the edge effect and antenna effect in THz s-SNOM. Our results indicate that the proposed model can give us a better understanding of the near-field interactions and direct the parameter design of the probe for THz s-SNOM. [ABSTRACT FROM AUTHOR]

The Calibration Microwave Radiometer (CMR) is a three-band radiometer deployed on the HY-2C satellite in a near-Earth orbit, and since it launched, there are few studies presented on the performance of CMR to date. Therefore, this paper focuses on providing an assessment of HY-2C CMR brightness temperature and wet troposphere correction (WTC). CMR works at 18.7 GHz, 23.8 GHz and 37 GHz in a nadir-viewing direction, aligned with the HY-2C radar altimeter. The wet troposphere path delay of the radar altimeter signal caused by water vapour and cloud liquid water content can be monitored and corrected by CMR. In this paper, guided by the concept of antenna pattern correction algorithm and a purely statistical method, we directly establish the function between the CMR antenna temperature and the model-derived WTC calculated by the European Centre from Medium-Range Weather Forecasting (ECMWF) Reanalysis data, which can obtain the brightness temperature and the WTC of CMR simultaneously. Firstly, the algorithm principle of CMR to establish the function between the antenna temperature and the model-derived WTC is introduced, and then the brightness temperature of CMR is evaluated using reference brightness temperatures of the Advanced Microwave Radiometer 2 (AMR-2) on Jason-3 satellite at crossover points. Furthermore, the performance of the CMR WTC is validated in three ways: (1) directly comparing with the colocated WTC measured by Jason-3 AMR-2, (2) directly comparing with model-derived WTC from ECMWF, which allows a rapid check at a global scale, (3) comparing the standard deviation of the Sea Surface Height (SSH) difference at crossover points using different WTC retrieval methods. The linear fit with Jason-3 brightness temperature and WTC in all non-precipitation conditions demonstrated a good agreement with Jason-3. In addition, the WTC of CMR has an obvious decrease in the standard deviation of the SSH difference compared with model-derived WTC, indicating the CMR can significantly improve the accuracy of the HY-2C SSH measurements. All the assessments indicate that the CMR performances are satisfying the expectations and fulfilling the mission requirements. [ABSTRACT FROM AUTHOR]

Straight bevel gears are widely used in mining equipment, ships, heavy industrial equipment, and other fields due to their high capacity and robust transmission. Accurate measurements are essential in order to determine the quality of bevel gears. We propose a method for measuring the accuracy of the top surface profile of the straight bevel gear teeth based on binocular visual technology, computer graphics, error theory, and statistical calculations. In our method, multiple measurement circles are established at equal intervals from the small end of the top surface of the gear tooth to the large end, and the coordinates of the intersection points of these circles with the tooth top edge lines of the gear teeth are extracted. The coordinates of these intersections are fitted to the top surface of the tooth based on NURBS surface theory. The surface profile error between the fitted top surface of the tooth and the designed surface is measured and determined based on the product use requirements, and if this is less than a given threshold, the product is acceptable. With a module of 5 and an eight-level precision, such as the straight bevel gear, the minimum surface profile error measured was −0.0026 mm. These results demonstrate that our method can be used to measure surface profile errors in the straight bevel gears, which will broaden the field of in-depth measurements for the straight bevel gears. [ABSTRACT FROM AUTHOR]

Proper selections of microwave-transmitting thermal control materials play key roles in the orbit operations of aerospace microwave detectors. Currently, white paint and germanium-coated polyimide film are common choices for microwave-transmitting thermal control, which have been extensively studied from the perspective of material compositions, ignoring the influence of the micro-nano structure on the surface of the material. Inspired by Saharan silver ants relying on micro-nano structures of hairs for heat protection and dissipation, in this paper, based on finite element simulation, spectral characteristics (e.g., solar absorptance and infrared emissivity) of three types of biomimetic micro-nano structures (e.g., including pyramid, tetrahedron, and triangular prism) were studied and compared. Simulation results revealed that spectral characteristics of these biomimetic micro-nano structures were mainly regulated by the tip height and the air gap; the decrease of the tip height led to the decrease in the solar absorptance and the infrared emissivity; the solar absorptance was decreased, and the infrared emissivity was increased with the increase of the air gap. When these biomimetic structures were compared with flat surfaces without micro-nano topographies, a decrease of the solar absorptance to block the incidence of heat flow from the sun and an increase of the infrared emissivity to dissipate its own heat to the space were located, which may give suggestions on the adjustment of spectral characteristics of aerospace microwave-transmitting thermal control materials. [ABSTRACT FROM AUTHOR]

In clinical diagnosis, radiological reports are essential to guide the patient's treatment. However, writing radiology reports is a critical and time-consuming task for radiologists. Existing deep learning methods often ignore the interplay between medical findings, which may be a bottleneck limiting the quality of generated radiology reports. Our paper focuses on the automatic generation of medical reports from input chest X-ray images. In this work, we mine the associations between medical discoveries in the given texts and construct a knowledge graph based on the associations between medical discoveries. The patient's chest X-ray image and clinical history file were used as input to extract the image–text hybrid features. Then, this feature is used as the input of the adjacency matrix of the knowledge graph, and the graph neural network is used to aggregate and transfer the information between each node to generate the situational representation of the disease with prior knowledge. These disease situational representations with prior knowledge are fed into the generator for self-supervised learning to generate radiology reports. We evaluate the performance of the proposed method using metrics from natural language generation and clinical efficacy on two public datasets. Our experiments show that our method outperforms state-of-the-art methods with the help of a knowledge graph constituted by prior knowledge of the patient. [ABSTRACT FROM AUTHOR]

Ceramic adhesive structures have been increasingly used in aerospace applications. However, the peaks of the signal on the upper and lower surface of the adhesive layer are difficult to measure directly due to the thin thickness of the adhesive layer and the effect of the attenuation dispersion of the ceramic layer. Thus, the existing non-destructive testing techniques have been ineffective in detecting adhesive quality. In this paper, the thickness of the adhesive layer is measured using terahertz time-domain spectroscopy. A sparse deconvolution method is proposed for the terahertz time-domain spectral signal of ceramic adhesive structures with different adhesive layer thicknesses. The results show that the methods proposed in this paper can realize the separation of reflection signals for glue layers with a thickness of 0.20 mm. By comparing with a wavelet denoising method and a modified covariance method (AR/MCM), the effectiveness of the sparse deconvolution method in estimating the thickness of the glue layer is demonstrated. This work will provide the theoretical and experimental basis for using terahertz time-domain spectroscopy to detect the homogeneity of ceramic adhesive structures. [ABSTRACT FROM AUTHOR]

Traditional GaAs-based frequency multipliers still exhibit great challenges to meet the demand for solid-state high-power THz sources due to low breakdown voltage and heat dissipation of the Schottky barrier diode (SBD). In this study, a GaN SBD chain was fabricated with n−/n+-GaN structure. As a consequence, the breakdown voltage of 54.9 V at 1 μA and cut-off frequency of 587.5 GHz at zero bias were obtained. A 120 GHz frequency-doubler module based on the GaN SBD chain was designed and fabricated. When driven with 500 mW input power in a continuous wave, the output power of the frequency-doubler module was 15.1 mW at 120 GHz. Moreover, the experiments show that the frequency-doubler module can endure an input power of 2 W. In addition, it is worth noting that the SBD chain works well at an anode temperature of 337.2 °C. [ABSTRACT FROM AUTHOR]

In recent years, many researchers have devoted time to designing algorithms used to introduce external information from knowledge graphs, to solve the problems of data sparseness and the cold start, and thus improve the performance of recommendation systems. Inspired by these studies, we proposed KANR, a knowledge graph-enhanced attention aggregation network for making recommendations. This is an end-to-end deep learning model using knowledge graph embedding to enhance the attention aggregation network for making recommendations. It consists of three main parts. The first is the attention aggregation network, which collect the user's interaction history and captures the user's preference for each item. The second is the knowledge graph-embedded model, which aims to integrate the knowledge. The semantic information of the nodes and edges in the graph is mapped to the low-dimensional vector space. The final part is the information interaction unit, which is used for fusing the features of two vectors. Experiments showed that our model achieved a stable improvement compared to the baseline model in making recommendations for movies, books, and music. [ABSTRACT FROM AUTHOR]

In recent years, the research of combining a knowledge graph with recommendation systems has caused widespread concern. By studying the interconnections in knowledge graphs, potential connections between users and items can be discovered, which provides abundant and complementary information for recommendation of items. However, most existing studies have not effectively established the relation between entities and users. Therefore, the recommendation results may be affected by some unrelated entities. In this paper, we propose a neighborhood aggregation collaborative filtering (NACF) based on knowledge graph. It uses the knowledge graph to spread and extract the user's potential interest, and iteratively injects them into the user features with attentional deviation. We conducted a large number of experiments on three public datasets; we verifyied that NACF is ahead of the most advanced models in top-k recommendation and click-through rate (CTR) prediction. [ABSTRACT FROM AUTHOR]

On the basis of the W-band power source, a single-stage frequency quadrupler method was used to implement two 335 GHz frequency quadruplers. The two frequency quadruplers adopted a traditional binomial matching structure and an improved gradient line matching structure, respectively. An idle loop was added to the overall circuit in the design of the DC filter and low-pass filter. The improved gradient line matching structure reduced the circuit length while increasing the bandwidth, effectively reducing the power loss on the transmission line. A micro-strip circuit was fabricated with a 50 μm thick quartz circuit and was mounted onto a split waveguide block. The results showed that the output power of the quadrupler with the improved matching structure was better than that of the quadrupler with the conventional matching structure. The peak output power of the improved frequency quadrupler was 4.75 mW at 333 GHz when driven with 200 mW. In contrast, this improved structure broadened the bandwidth by 8 GHz and reduced the length of the substrate by 0.607 mm, effectively reducing the length of the traditionally designed circuit by 11.5%. [ABSTRACT FROM AUTHOR]

The development of a millimeter-wave unbalanced frequency tripler based on the nonlinear characteristics of planar Schottky varactors is presented. The entire module is designed by hybrid integration. A frequency multiplier circuit model was established to reflect the influence of diode parameters and the impedance matching on the multiplier in different frequency bands. The effect of junction imbalance on the output power of the frequency multiplier was investigated and the multiplier was improved based on the basic design. The addition of a cut microstrip stub in the improved diode unit reduced the impact of a power imbalance on frequency multiplier performance. The characteristics of the multiplier circuit were analyzed by the full-wave electromagnetic simulation of the three-dimensional structure and the harmonic balance simulation of the circuit. Test results showed that the peak output power of the improved frequency tripler was 12.6 mW at 277 GHz with an input power of 200 mW, an effective 12% improvement over the basic design. [ABSTRACT FROM AUTHOR]

Based on deep neural network multi-task learning technology, lane image detection is studied to improve the application level of driverless technology, improve assisted driving technology and reduce traffic accidents. The lane line database published by Caltech and Tucson company is used to extract the ROI (Region of Interest), scale, and inverse perspective transformation as well as to preprocess the image, so as to enrich the data set and improve the efficiency of the algorithm. In this study, ZFNet is used to replace the basic networks of VPGNet, and their structures are changed to improve the detection efficiency. Multi-label classification, grid box regression and object mask are used as three task modules to build a multi-task learning network named ZF-VPGNet. Considering that neural networks will be combined with embedded systems in the future, the network will be compressed to CZF-VPGNet without excessively affecting the accuracy. Experimental results show that the vision system of driverless technology in this study achieved good test results. In the case of fuzzy lane line and missing lane line mark, the improved algorithm can still detect and obtain the correct results, and achieves high accuracy and robustness. CZF-VPGNet can achieve high real-time performance (26FPS), and a single forward pass takes about 36 ms or less. [ABSTRACT FROM AUTHOR]

The present work proposes a 180–225 GHz broadband frequency doubler monolithic microwave integrated circuit (MMIC) based on a novel Schottky barrier diode (SBD) terminal structure denoted as a Schottky metal-brim (SMB). Compared with an MMIC adopting the conventional SBD terminal structure, preliminary measurements show that the maximum output power of the MMIC adopting the SMB structure increases from 0.216 mW at 206 GHz to 0.914 mW at 208 GHz. Analysis of the nonlinear current–voltage and capacitance–voltage characteristics of the two terminal structures based on an extended one-dimensional drift-diffusion model, indicates that the SMB structure provides significantly better conversion efficiency than the conventional SBD structure by eliminating the accumulation of charge and additional current paths near the Schottky electrode edge. It provides a feasible scheme for the optimization of MMIC applications requiring high power and high efficiency. [ABSTRACT FROM AUTHOR]

A global design method for a terahertz monolithic integrated frequency multiplier is proposed. Compared with a traditional independent design, the method in this paper adopts overall optimization and combines the device with the circuit design. The advantage is that it provides a customized design for frequency multipliers according to specifications. On the basis of the gallium arsenide process of the Institute of Microelectronics, Chinese Academy of Sciences, two types of Schottky diodes have been developed to meet the needs of different designs. On the one hand, a Schottky diode with a 3 μm junction's diameter was used in the design of the 200 GHz balanced doubler and, on the other hand, a diode with a 5 μm diameter was used in the 215 GHz unbalanced tripler. The measured results indicated that the output power of the doubler was more than 250 μW at 180~218 GHz, and the maximum was 950 μW at 198 GHz when driven with 12.3 mW, whereas that of the tripler was above 5 mW at 210~218 GHz and the maximum exceeded 10 mW. Such frequency multiplier sources could be widely used in terahertz imaging, radiometers, and so on. [ABSTRACT FROM AUTHOR]

With the arrival of 5G networks, cellular networks are moving in the direction of diversified, broadband, integrated, and intelligent networks. At the same time, the popularity of various smart terminals has led to an explosive growth in cellular traffic. Accurate network traffic prediction has become an important part of cellular network intelligence. In this context, this paper proposes a deep learning method for space-time modeling and prediction of cellular network communication traffic. First, we analyze the temporal and spatial characteristics of cellular network traffic from Telecom Italia. On this basis, we propose a hybrid spatiotemporal network (HSTNet), which is a deep learning method that uses convolutional neural networks to capture the spatiotemporal characteristics of communication traffic. This work adds deformable convolution to the convolution model to improve predictive performance. The time attribute is introduced as auxiliary information. An attention mechanism based on historical data for weight adjustment is proposed to improve the robustness of the module. We use the dataset of Telecom Italia to evaluate the performance of the proposed model. Experimental results show that compared with the existing statistics methods and machine learning algorithms, HSTNet significantly improved the prediction accuracy based on MAE and RMSE. [ABSTRACT FROM AUTHOR]

This paper proposes a novel sub-harmonic mixing topology. Based on the proposed topology and the precise three-dimensional electromagnetic model of the Schottky barrier diode; a novel 183 GHz solid-state sub-harmonic mixer is designed and measured. By adding a compact low-pass filter near the ground of the mixer's circuit, the effect on the mixer's RF performance of the random error of the conductive adhesive in assembling is effectively decreased. The test results show that the optimal single-sideband conversion loss of the mixer is 8.1dB@183GHz when the local oscillator signal is 4mw@91GHz. In the RF bandwidth from 173 GHz to 191 GHz, the single-sideband conversion loss is less than −10.6 dB. At the same time, the RF port return loss is less than 9.8 dB. [ABSTRACT FROM AUTHOR]

To effectively reduce the requirement of Local Oscillator (LO) power, this paper presents the design and measurement of a biased sub-harmonic mixer working at the center frequency of 0.67 THz in hybrid integration. Two discrete Schottky diodes were placed across the LO waveguide in anti-series configuration on a 50 μm thick quartz-glass substrate, and chip capacitors were not required. At the driven of 3 mW@335 GHz and 0.35 V, the mixer had a minimum measured Signal Side-Band (SSB) conversion loss of 15.3 dB at the frequency of 667 GHz. The typical conversion loss is 18.2 dB in the band of 650 GHz to 690 GHz. [ABSTRACT FROM AUTHOR]

Based on a W-band high-power source, two schemes are proposed to realize a 335 GHz frequency multiplier source. The first scheme involves producing a 335 GHz signal with a two-stage doubler. The first doubler adopts two-way power-combined technology and the second stage is a 335 GHz doubler using a balanced circuit to suppress the odd harmonics. The measured output power was about 17.9 and 1.5 dBm at 167.5 and 335 GHz, respectively. The other scheme involves producing a 335 GHz signal with a single-stage quadrupler built on 50 µm thick quartz circuit adopting an unbalanced structure. The advantage of the unbalanced structure is that it can provide bias to the diodes without an on-chip capacitor, which is hard to realize with discrete devices. The measured output power was about 5.8 dBm at 337 GHz when driven with 22.9 dBm. Such 335 GHz frequency multiplier sources are widely used in terahertz imaging, radiometers, and so on. [ABSTRACT FROM AUTHOR]

The rapid change of topology is one of the most important factors affecting the performance of the routing protocols of flying ad hoc networks (FANETs). A routing scheme suitable for highly dynamic mobile ad hoc networks is proposed for the rapid change of topology in complex scenarios. In the scheme moving nodes sense changes of the surrounding network topology periodically, and the current mobile scenario is confirmed according to the perceived result. Furthermore, a suitable routing protocol is selected for maintaining network performances at a high level. The concerned performance metrics are packet delivery ratio, network throughput, average end-to-end delay and average jitter. The experiments combine the random waypoint model, the reference point group mobility model and the pursue model to a chain scenario, and simulate the large changes of the network topology. Results show that an appropriate routing scheme can adapt to rapid changes in network topology and effectively improve network performance. [ABSTRACT FROM AUTHOR]

On the basis of the W-band power source, a single-stage frequency quadrupler method was used to implement two 335 GHz frequency quadruplers. The two frequency quadruplers adopted a traditional binomial matching structure and an improved gradient line matching structure, respectively. An idle loop was added to the overall circuit in the design of the DC filter and low-pass filter. The improved gradient line matching structure reduced the circuit length while increasing the bandwidth, effectively reducing the power loss on the transmission line. A micro-strip circuit was fabricated with a 50 μm thick quartz circuit and was mounted onto a split waveguide block. The results showed that the output power of the quadrupler with the improved matching structure was better than that of the quadrupler with the conventional matching structure. The peak output power of the improved frequency quadrupler was 4.75 mW at 333 GHz when driven with 200 mW. In contrast, this improved structure broadened the bandwidth by 8 GHz and reduced the length of the substrate by 0.607 mm, effectively reducing the length of the traditionally designed circuit by 11.5%.