Table of Content

15 December 2020 Volume 33 Issue 12

  

ModelingAnd Simulation of Primary Winding Piecewise PMSLM System Based on Vector Control

FAN Lele,WANG Xudong,FENG Haichao,XU Xiaozhuo

Electronic Science and Technology. 2020, 33(12):  1-6.  doi:10.16180/j.cnki.issn1007-7820.2020.12.001

Abstract ( 420 )   HTML ( 19 )   PDF (860KB) ( 70 )  

Figures and Tables | References | Related Articles | Metrics

Aiming at the special sectional power supply problem of primary winding segment permanent magnet linear synchronous motor, the mathematical model of primary winding segment permanent magnet linear synchronous motor was constructed by space vector decoupling method. In addition, on this basis, a vector control model for primary winding segment power supply was established. By comparing the resultant force produced by the primary winding of each element , the thrust produced by the coupling of the moving body and the primary winding of a single segment, the accuracy of the model was judged. The simulation experiments were carried out under the environment of MATLAB/Simulink. The results showed that the thrust of the moving body was equal between segments and within segments. Therefore, the model built could reflect the displacement of the mover effectively and quickly, realized precise winding switching and achieved the purpose of sectional power supply.

Simulation and Analysis of HF Communication Protocol Based on OPNET

XUE Huanjie,CHEN Xiaoli,GUAN Boran

Electronic Science and Technology. 2020, 33(12):  7-11.  doi:10.16180/j.cnki.issn1007-7820.2020.12.002

Abstract ( 447 )   HTML ( 10 )   PDF (1292KB) ( 45 )  

Figures and Tables | References | Related Articles | Metrics

Because of the undestructibility of ionospheric channel, it is very important to study short-wave communication protocol in the field of emergency communication. Based on the third-generation short-wave automatic link protocol, this paper studies and analyzes the multi-access technology with the new technology. The feasibility of the scheme is verified using OPNET Modeler 14.5 network simulation platform. In the short-wave communication network model, the pipeline mechanism uses the long-term channel prediction software VOACAP to obtain the ionospheric channel quality parameters for wireless channel modeling. The core function and C programming language are adopted to establish the 3g-ale node model and the related process model and network model. Simulation experiments show that the channel quality has the most significant effect on the successful ratio of the chain between network nodes, and the number of network nodes and the distance between two nodes have the lower effect.

A Domain Adaptive Depth Estimation Method for Structural Perception Loss

ZHAN Yan,ZHANG Juan

Electronic Science and Technology. 2020, 33(12):  12-16.  doi:10.16180/j.cnki.issn1007-7820.2020.12.003

Abstract ( 217 )   HTML ( 5 )   PDF (1602KB) ( 32 )  

Figures and Tables | References | Related Articles | Metrics

In the domain adaptive depth estimation method, the structural differences between domain images are large. In view of this problem, a domain adaptive depth estimation method for structural perceptual loss is proposed. This method uses pre-trained convolutional neural networks to extract features from images, and measures structural similarity on features, which reduces the difference between domain images and improved the stability of the transform module. This method uses synthetic image depth pairs and real image training, and eliminates the requirement of depth labels and physical geometric information for real images. Experiments on the KITTI dataset achieve a depth accuracy rate of 96.6%, which proves that the method can effectively improve the depth accuracy.

Research on Dynamic Gesture Recognition Based on Dense Trajectories Features

WANG Hehe,LI Feifei,CHEN Qiu

Electronic Science and Technology. 2020, 33(12):  17-21.  doi:10.16180/j.cnki.issn1007-7820.2020.12.004

Abstract ( 289 )   HTML ( 10 )   PDF (1566KB) ( 28 )  

Figures and Tables | References | Related Articles | Metrics

In view of at the low performance of feature point detection for gesture recognition, the paper propose a new edge feature point detection method. By using the feature description method of the dense trajectories method, the method of support vector machine classification learning is uses to realize dynamic gesture recognition. This method increases the number of edge trajectories effectively, which brings benefits to the final recognition result. The method is evaluated in Cambridge University Gesture Dataset and Sheffield Gesture Dataset, and 99.11% and 99.72% gesture recognition accuracy is obtained respectively, which embodies the excellent performance of the algorithm in the above datasets.

Improved Background Subtraction Based on Feature Fusion

GUAN Hongyun,SU Zhentao,WANG Chen

Electronic Science and Technology. 2020, 33(12):  22-27.  doi:10.16180/j.cnki.issn1007-7820.2020.12.005

Abstract ( 308 )   HTML ( 16 )   PDF (1282KB) ( 71 )  

Figures and Tables | References | Related Articles | Metrics

The background difference method has higher reliability in completely and quickly segmenting the target image, but the detection effect is not good in the case of background disturbance and illumination variation. This study proposes an improved background difference algorithm based on feature fusion. This algorithm fuses spatio-temporal local binary pattern texture features and color features, and consideres the confidence and similarity scores of the two features to obtain the background probability, following by foreground segmentation. The background pixels are used for background template updates, which is designed to better solve the problem of target detection in complex backgrounds. The experimental results show that the detection effect of this algorithm is better than other similar algorithms. While maintaining the robustness and complexity of the background difference algorithm, it displayes good detection effect under the background disturbance and illumination changes.

Network Security Situation Assessment Based on Multimodal Feature Fusion

LI Kang

Electronic Science and Technology. 2020, 33(12):  28-31.  doi:10.16180/j.cnki.issn1007-7820.2020.12.006

Abstract ( 822 )   HTML ( 30 )   PDF (662KB) ( 169 )  

Figures and Tables | References | Related Articles | Metrics

In view of the problems of single information source and low early warning quality of network security monitoring equipment, a network security situation assessment method integrating multiple data sources is proposed in this study. The NSSA framework of network security situation is constructed by introducing Endsley model and agent theory. By using the idea of radial basis function neural network, the fusion of multi-source heterogeneous data is realized through eliminating redundant noise and irrelevant signals. Therefore, a network security situation assessment method with multi-modal feature fusion is proposed. The simulation results of MATLAB software show that compared with the traditional BP and RBF neural networks, the proposed network security situation assessment method has better learning ability and generalization ability.

The Design of Wideband Tunable Metamaterial Absorber Based on Electrorheological Fluid

WANG Liansheng,XIA Dongyan,FU Quanhong,DING Xueyong,WANG Yuan

Electronic Science and Technology. 2020, 33(12):  32-37.  doi:10.16180/j.cnki.issn1007-7820.2020.12.007

Abstract ( 266 )   HTML ( 5 )   PDF (1900KB) ( 31 )  

Figures and Tables | References | Related Articles | Metrics

In order to better apply the metamaterial absorber to the practice, a new kind of wideband tunable metamaterial absorber based on electrorheological fluid is proposed in this paper. The wideband and tunable absorption are achieved by loading resistor and electrorheological fluid with electrically tunable permittivity in metamaterial absorber. The simulation results showed that the absorption of absorber reaches up to 80% from 8.296 GHz to 15.128 GHz, and reache up to 90% from 11.5 GHz to 15 GHz, which realizes the property of wideband absorption. The absorption band gradually shifts to lower frequency with the increasing of electric field strength applied on electrorheological fluid, realizing the tunable property of wideband absorption. Moreover, the absorption property of absorber is insensitive to polarization due to its rotational symmetry.

Adaptive Sliding Mode Control for Nonlinear Systems with Mismatched Quantization

ZHANG Lele,SU Qianmin

Electronic Science and Technology. 2020, 33(12):  38-43.  doi:10.16180/j.cnki.issn1007-7820.2020.12.008

Abstract ( 224 )   HTML ( 5 )   PDF (721KB) ( 28 )  

Figures and Tables | References | Related Articles | Metrics

This study aims to investigate the asymptotic stability problem of nonlinear systems. The encoder and decoder parameters of the quantizer may be mismatched, and the unmatched quantization parameters may further increase the complexity and difficulty of the controller design. In this paper, an adaptive sliding film control method is designed by establishing a time-varying proportional model of quantitative parameters, combining observer-based techniques and sliding mode control techniques. The proposed method completely eliminates the influence of mismatched quantization parameters, nonlinearity and external disturbance, and realizes the asymptotic stability of the closed-loop system and drives the state trajectory of systems to the sliding mode surface.Finally, the effectiveness of the control strategy is verified by simulation experiments.

Research Progress in Preparation of Graphene by Mechanical Shear Stripping

WANG Jiatao,WEI Jingtao,QIAN Jie,WU Zhangyong,WANG Tingyou

Electronic Science and Technology. 2020, 33(12):  44-48.  doi:10.16180/j.cnki.issn1007-7820.2020.12.009

Abstract ( 888 )   HTML ( 26 )   PDF (940KB) ( 148 )  

Figures and Tables | References | Related Articles | Metrics

The excellent electrical, optical and mechanical properties of graphene have set off a research boom in various fields, and its preparation method has become the focus of research. Mechanical shear stripping method has attracted much attention because of its low cost, simple operation, and no damage to the intrinsic properties of graphene. This paper introduces the properties and stripping mechanism of graphene, and reviewes the low energy pure shearing method, the three-stick grinding and stripping method and the ball milling method in the mechanical peeling method, and compares their respective characteristics and problems. Since the ball milling method is currently the most promising method for producing graphene on a large scale. the ball milling process is optimized in this paper, and a grinding disc structure ball mill is proposed. The ball mill effectively avoids the impact force between the grinding balls, protects the graphite lattice from being broken, and improves the shearing efficiency.

Image Splicing Detection Based on Optimal Color Channel

XIONG Shiting,ZHANG Yujin,LIU Tingting,FANG Xiangyu

Electronic Science and Technology. 2020, 33(12):  49-53.  doi:10.16180/j.cnki.issn1007-7820.2020.12.010

Abstract ( 264 )   HTML ( 4 )   PDF (2001KB) ( 30 )  

Figures and Tables | References | Related Articles | Metrics

In view of the problem that different color channels have effect on the noise estimation value, an image splicing detection method based on optimal color channel is proposed. The noise on the optimal color channel is estimated by principal component analysis and the method of K-means clustering is used to cluster according to the noise value. The clustering result is divided into suspicious part and non-suspicious part. The two-phase strategy can further locate the tampering area. The method is effective when the noise value difference between the original area and the splicing area was large or small, and the splicing area can be located. Experiments show that compared with the existing methods, the proposed method achieves good detection results, and the performance is better.

Scene Recognition Algorithm Based on Convolutional Neural Networks and Multi-Scale space Encoding

MIAO Ran,LI Feifei,CHEN Qiu

Electronic Science and Technology. 2020, 33(12):  54-58.  doi:10.16180/j.cnki.issn1007-7820.2020.12.011

Abstract ( 258 )   HTML ( 11 )   PDF (1647KB) ( 41 )  

Figures and Tables | References | Related Articles | Metrics

A scene image is generally composed of some foreground objects and background contexts with a certain spatial layout. Due to different scales, viewpoints, and backgrounds, there exists large intra-class variation within the same scene class. On the other hand, the common objects also result in a certain inter-class similarities among heterogeneous scenes as well. Consequently, the multi-scale space encoding based on convolutional neural networks (CNN) for scene representation is proposed in the study, which combines multi-scale dense sampling method, CNN algorithm, and multi-scale space encoding method. The multi-scale encoding method spatially partitions the sampling grid many times, and then aggregates the CNN features within sub-regions with different shapes for generating the multi-scale space VLAD. The experiment is carried out on the Scene15 scene dataset, and the test results show that the test accuracy reaches 94.67%.

Research on Complex Background Image Classification Method Based on Deep Learning

CHENG Junhua,ZENG Guohui,LIU Jin

Electronic Science and Technology. 2020, 33(12):  59-66.  doi:10.16180/j.cnki.issn1007-7820.2020.12.012

Abstract ( 1144 )   HTML ( 32 )   PDF (2424KB) ( 191 )  

Figures and Tables | References | Related Articles | Metrics

To solve the problem that the complex background image cannot be easily recognized by background interference, a convolutional neural network image classification method based on foreground region segmentation mechanism is proposed. The foreground region of image is automatically segmented by using the full convolutional neural network, and is located by the minimum bounding box around it. A convolutional neural network is constructed to distinguish different foreground region categories, thereby realizing the classification of the complex background image. The proposed method is used to perform contrast experiments on simply and complex background image datasets extracted from the public dataset. Some effective methods, such as transfer learning and data augmentation, are used for model optimization. The experimental results show that the proposed method has higher accuracy than only classification CNN on both datasets, and the extent of model accuracy improvement on the complex background image is much larger than that on the simple background image. These results prove that the introduction of foreground region segmentation can improve the accuracy of complex background image classification model, highlight the characteriskics of the foreground region and reduce the interference of background factors.

Parameters Selection for LSSVM Based on Artificial Fish Swarm-Shuffled Frog Jump Algorithms Optimization in Short-Term Load Forecasting

YANG Haizhu,JIANG Zhaoyang,LI Menglong,KANG Le

Electronic Science and Technology. 2020, 33(12):  67-74.  doi:10.16180/j.cnki.issn1007-7820.2020.12.013

Abstract ( 234 )   HTML ( 6 )   PDF (1107KB) ( 41 )  

Figures and Tables | References | Related Articles | Metrics

Short-term load forecasting plays a key role in safe dispatching and economic operation of power system.The parameters of the LSSVM directly affect the prediction effect during the load forecasting accuracy. In order to improve LSSVM load prediction accuracy, a method based on levy adaptive vision artificial fish swarm-shuffled frog leaping algorithm for parameter optimization of LSSVM is proposed. LSSVM is trained by historical data such as load and weather in a certain area. The LAFSA-SFLA-LSSVM forecasting model, the LAVAFSA-SFLA-LSSVM forecasting model and the AFSA-LSSVM forecasting model are established for power load forecasting in a certain area within 24 hours of a certain day. The results show that the accuracy of the LAVAFSA-SFLA-LSSVM forecasting model is higher than the AFSA-LSSVM forecasting model and the LAFSA-SFLA-LSSVM forecasting model, and the prediction error is smaller.

Design of English-Chinese Translation System Based on Variational Model

ZHENG Meng

Electronic Science and Technology. 2020, 33(12):  75-78.  doi:10.16180/j.cnki.issn1007-7820.2020.12.014

Abstract ( 247 )   HTML ( 4 )   PDF (575KB) ( 28 )  

Figures and Tables | References | Related Articles | Metrics

Anti-neural machine translation method is a hot machine translation algorithm at present, but the translation accuracy of traditional anti-neural network model depends on a large number of corpus data sets, and the model training takes a lot of time. When the corpus is scarce, the model translation quality is poor. Aiming at the shortcomings of traditional anti-neural network machine translation algorithm, this paper combines variational algorithm with anti-neural network to train corpus data. Experimental results show that the BLEU value of the variational anti-neural network translation is obviously improved compared with the traditional translation algorithm.When the number of training corpus is scarce, the BLEU value of the model is greatly improved compared with other algorithms, which shows that the proposed algorithm model can effectively shorten the training time of data, elevate the training accuracy of data and improve the translation quality of sentences.