Table of Content

25 January 2020, Volume 48 Issue 1

Previous Issue    Next Issue

Mechanical Engineering

PCA Feature Frequency Extraction Algorithm Based on SVD Principle and Its Application



GUO Mingjun, LI Weiguang, YANG Qijiang, et al

2020, 48(1):  1-9.  doi:10.12141/j.issn.1000-565X.190103

Asbtract ( 553 )   HTML ( 46)   PDF (2511KB) ( 128 )  

Related Articles | Metrics

A PCA feature frequency extraction algorithm based on SVD principle was proposed to solve the noise pollution problem in the measured displacement signal of rotor． Firstly，the intrinsic relationship between PCA and SVD was deduced theoretically． That is，the eigenvalue of the covariance matrix generated by PCA was equal to the square of the singular value of the matrix generated by SVD，and the eigenvector generated by PCA was equal to the left singular vector generated by SVD． Then，based on the above conclusions，a PCA feature frequency ex- traction algorithm based on SVD principle was proposed，and the effectiveness of the algorithm was verified by simu- lation signals． Finally，the algorithm was applied to purify axis orbits of the rotor of a large sliding bearing test bed． The axis orbits are clear and concentrated，the misalignment and friction faults were identified successfully．

Force Characteristics of Active Dynamic Vibration Absorber Based on Magnetic Actuator



FU Tao, YIN Zhihong, REN Yan, et al

2020, 48(1):  10-18.  doi:10.12141/j.issn.1000-565X.190313

Asbtract ( 268 )   HTML ( 7)   PDF (1007KB) ( 338 )  

Related Articles | Metrics

A new electromagnetic active vibration absorber was desiggned by using an electromagnet as the actua- tor，and its force characteristics was studied． Firstly，the mathematical model for electromagnetic force of actuator was built． Secondly，the kinematical equation of the active vibration absorber was deduced based on the mathe- matical model，and the characteristics of electromagnetic force and driving force was analyzed． Finally，conside- ring the inevitable variance between the practical driving force and the ideal driving force of active vibration absor- ber，the influence of the variance in amplitude，phase and frequency on the performance of the active vibration ab- sorber was analyzed analytically． Experiment bench for a plate was set up，and the theoretical analysis has been verified by experiments． The results show that the greater the amplitude and phase variance between the practical driving force and the ideal driving force is，the poorer the damping effect is; when there is frequency variance，ac- tive vibration absorber has no damping effect．

Design of Balance Flow Channel in T-type Extrusion Dies



MA Xiangjun

2020, 48(1):  19-24.  doi:10.12141/j.issn.1000-565X.190361

Asbtract ( 580 )   HTML ( 11)   PDF (607KB) ( 101 )  

Related Articles | Metrics

T-type extrusion dies is a kind of widely-used slot die． According to the present design method for flow channel in die，increasing the radius of the manifold or reducing the thickness of the choke zone was generally a- dopted to improve the uniformity of the melt exit flow rate along die width direction． However，increasing the radi- us of manifold increases the residence time of the melt in die，and reducing the thickness of choke zone causes a sharp increase of extrusion pressure，so other measures are demanded to meet requirement in practice． Two choke zones with different thicknesses were adopted to solve the problem． Under the precondition that the melt exit flow rate is uniform along the die width direction，the demarcation carve between the two choke zones were deduced based on rheology． In addition，the traditional T-type dies designing method was compared with the method pro- posed． The results show that the residence time of the melt can be significantly reduced by using the proposed de- sign method and the flow channel designed through proper selection of radius of manifold，the thickness of the two choke zones and demarcation curve between the two choke zones． ，and the extrusion pressure can be adjusted ac- cording to the processing requirements． Therefore，the proposed design method can be used to guild the design of flow channel in T-type dies．

Similarity Model for Lubrication Experiment of Spiral Bevel Gear and Influencing Parameters



WANG Yanzhong, YANG Kai, QI Ronghua, et al

2020, 48(1):  25-31.  doi:10.12141/j.issn.1000-565X.190310

Asbtract ( 325 )   HTML ( 9)   PDF (767KB) ( 89 )  

Related Articles | Metrics

The meshing process of the spiral bevel gear was simplified to the point contact between several curved surfaces and a plane． A similarity model subject to elastohydrodynamic lubrication was established by using the similarity theory． The model was then verified by the elastohydrodynamic lubrication theory． The influence of the different parameters on the system similarity was investigated． The analysis results show that under the premise of meeting the similar criteria，changing the entrainment speed and the viscosity of the lubricating oil has little influ- ence on the similarity degree; When the load，the pressure-viscosity coefficient，the radius of curvature and the e- lastic modulus are changed，the similarity error of the system becomes larger; the more the parameters change，the more obvious the error stacking effect is． Selecting the parameters reasonably according to the similarity model， the film thickness and its distribution in the meshing zone of the spiral bevel gear pair can be indirectly measured by the optical interferometry method． Finally，the experimental data were compared with the similarity model results． The result shows that when the entrainment speed is less than 1 m /s，the similarity model results are close to the experimental results，and the maximum error is 12. 82% ; when the entrainment speed changes greatly ，the simi- larity model results have large deviations．

Computer Science & Technology

Motion Deblurring Based on DeblurGAN and Low Rank Decomposition



SUN Jifeng ZHU Yating WANG Kai

2020, 48(1):  32-41,50.  doi:10.12141/j.issn.1000-565X.190038

Asbtract ( 550 )   HTML ( 9)   PDF (11080KB) ( 83 )  

Related Articles | Metrics

An end-to-end image motion deblurring method based on DeblurGAN was proposed with conditional generative adversarial network． In this method，the standard convolution layers in DeblurGAN were changed into bottleneck structures，and low-rank decomposition was further performed on the convolution layers in the bottleneck structures． Then two residual symmetric skip connections were added to accelerate the convergence of the network． In order to solve the problem that the restored images in the DeblurGAN are not clear，mutual information loss and the gradient image L1 loss were added to the network loss function． By maximizing the mutual information between the input image and its hidden feature，the extracted hidden feature can well represent the input information， thereby obtaining a clear restored image from the hidden feature，and the L1 loss helps to make the restored image with more significant edge． At the same time，the effectiveness of the proposed method was verified by experiments and compared with other existing similar algorithms． The results show that compared with DeblurGAN，the peak signal-to-noise ratio of the proposed method is higher; the structural similarity measure of the two methods is equiv- alent; the parameter quantity of our model is compressed to 3. 25% of DeblurGAN; the deblurring processing speed is increased by 3 times，and our model outperforms other existing similar algorithms．

Application of Binary Tree Model in Object Tracking



ZHENG Yunping LI Ruijun

2020, 48(1):  42-50.  doi:10.12141/j.issn.1000-565X.190173

Asbtract ( 517 )   HTML ( 10)   PDF (6505KB) ( 190 )  

Related Articles | Metrics

Object tracking has always been an important research topic in the field of computer vision． It is widely used in video surveillance，traffic monitoring，medical diagnosis and other fields． An object tracking algorithm based on binary tree model was proposed． The method divides the target area of the image into several homogene-ous blocks of different sizes，following the rule of the binary tree partition． The pixels in the block are similar and can be represented by a single value or vector whereas the pixels in different blocks differ from each significantly，thus forming the feature description model of the whole object． The CT algorithm，the quadtree model based algo-rithm (QT algorithm) and the proposed binary tree model based algorithm (BT algorithm) were compared from the aspects of accuracy and tracking speed． The results show that compared with the quadtree-based algorithm，the BT-based tracking algorithm can improve the tracking speed significantly without reducing the tracking accuracy．And compared with the discriminant CT-based algorithm，which is known for its fast tracking speed，the BT-based tracking accuracy is even better under the premise that the tracking speed is roughly equal．


Pavement Anomaly Detection Algorithm Based on High-order Dynamic Bayesian Network Embedding#br#



LI Bo ZHANG Honggang

2020, 48(1):  51-59.  doi:10.12141/j.issn.1000-565X.180583

Asbtract ( 418 )   HTML ( 8)   PDF (589KB) ( 174 )  

Related Articles | Metrics

Pavement anomalies can bring inconvenience to drivers and passengers，and even cause traffic acci-dents． A pavement anomaly detection algorithm based on sensor time series data was proposed． Considering the problem that sensing signals collected during driving are strongly high-order sequential correlative，high-order dy-namic Bayesian network classifier was constructed to realize the anomalies detection． Firstly，correlation analysis and Granger causality test were used to initialize the network structure． Secondly，the sensing signals were decom-posed by wavelet transform，and convolution neural network was used to realize network embedding． Finally，link prediction with minimal description length was used to optimize the network structure． The results show that，com-pared with the traditional method of time series classification，the proposed method can reduce fallout rate and missing rate，and increase F1 score on the sequential correlative signals，and thus is more robust．


Improved Stereo Matching Algorithm Based on PSMNet



LIU Jianguo, FENG Yunjian, JI Guo, et al

2020, 48(1):  60-69,83.  doi:10.12141/j.issn.1000-565X.190388

Asbtract ( 1075 )   HTML ( 21)   PDF (897KB) ( 270 )  

Related Articles | Metrics

Based on PSMNet stereo matching network，an improved stereo matching algorithm with shallow struc-ture and wide receptive field -SWNet was proposed，in order to solve the stereo matching problem in binocular vi-sion，reduce the number of parameters of the stereo matching network，reduce the computational complexity of the algorithm，and improve the practicability of the algorithm． The shallow structure means fewer layers，fewer pa-rameters and faster processing speed，while wide receptive field means that the network is more receptive and can acquire and retain more spatial information． SWNet consists of three parts: feature extraction，3D convolution and disparity regression． In the aspect of feature extraction，Atrous Spatial Pyramid Pool (ASPP) was introduced，which was used to extract multi-scale feature information． Feature fusion module was designed to fuse multi-scale feature information and build matching cost volume． The 3D convolutional neural network use the stack encoding
and decoding structure to further regularize the matching cost volume and obtain the corresponding relationship be-tween the feature points under different disparity conditions． Finally，the disparity map was obtained by regres-sion． SWNet performed well on both SceneFlow and KITTI 2015 public datasets，with a 48. 9% reduction in the number of parameters and a 2. 24% mismatching rate compared to the reference algorithm PSMNet．


Forest Fire Recognition Based on Color and Texture Features



LI Juhu FAN Ruixian CHEN Zhibo

2020, 48(1):  70-83.  doi:10.12141/j.issn.1000-565X.190181

Asbtract ( 807 )   HTML ( 15)   PDF (6794KB) ( 142 )  

Related Articles | Metrics

A flame recognition algorithm based on the partitioned LBP histogram feature combined with the LPQ histogram feature was proposed according to the unique color and texture feature of flame． The algorithm was de-signed to reduce the false positive rate of forest fire in the presence of flame-like interference source and increase the speed of fire warning． Firstly，the rule in YCbCr color space was used to detect the suspected flame region．Secondly，LBP and LPQ were used to extract the texture from the spatial domain and frequency domain． Then the feature vector was obtained by combining the extracted texture features． Finally，the feature vector was inputted into support vector machine (SVM) for flame recognition． The experimental results show that the algorithm is ro-bust and has a high detection rate． When there is a flame-like interference source，the accuracy of flame iden-tification of the test set can reach 94.55%． Compared with deep learning algorithm，the proposed algorithm can signifi-cantly improve the speed of fire warning while ensuring a high accuracy． Its forecasting time is 1/4 of the forecasting time of DBN，and 1/50 of that of CNN． Thus the algorithm provides a basis for fast and accurate forest fire warning．


Electronics, Communication & Automation Technology

Stereo Matching Algorithm Based on Cross-scale Random Walk



LI Qiang, DUAN Ziyang, ZHANG Yifan, et al

2020, 48(1):  84-92.  doi:10.12141/j.issn.1000-565X.190340

Asbtract ( 349 )   HTML ( 6)   PDF (6959KB) ( 70 )  

Related Articles | Metrics

Traditional stereo matching algorithms are mostly based on the correspondence between two image pixels or partial blocks，finding the disparity map at a single scale． But these algorithms can not model the correspon-dence between low-texture and repeated texture regions，resulting in a limited accuracy of the obtained disparity map． Considering the human visual system processes the received visual signals on different scales，a cross-scale restart and random walk algorithm was proposed to improve the above problems． Firstly，the matching cost of the scene images was calculated． Then using super pixel segmentation for rapid initial aggregation，and using the re-start and random walk algorithm to optimize it globally． Finally，an effective fusion update of the matching cost was realized by adopting the cross-scale model，and then a disparity map of the scene image was obtained． The ex-perimental results on the Middlebury dataset show that，compared with the traditional cross-scale stereo matching algorithm，the proposed algorithm can effectively reduce the average mismatch rate of the scene image in all regions and non-occluse regions by 1 percentage and 3 percentage points respectively，and obtain high-precision disparity
map．


Deception Mechanism of FDA Jammer to Passive Radar Interferometer Direction Finding System#br#



WANG Bo, XIE Junwei, GE Jiaang, et al

2020, 48(1):  93-103.  doi:10.12141/j.issn.1000-565X.190264

Asbtract ( 389 )   HTML ( 3)   PDF (1093KB) ( 284 )  

Related Articles | Metrics

In the radar electronic warfare，the enemy's passive detection radar can realize the direction finding and positioning by receiving the jammer signal，threatening the safety of our jammer． Thus a direction-based cross-lo-cation spoofing technique based on Frequency Divers Array (FDA) was put forward． Firstly，the FDA array fac-tor and phase pattern using nonlinear frequency offset were obtained by Euler's formula based on the establishment of the FDA array model． Then the FDA array was used to measure the signal arrival angle β and position x of the virtual transmitter in x-axis of the coordinate． Next，based on the effect of the interferometer's direction-finding de-ception，the deception principle and positioning accuracy of the FDA jammer to the direction-finding cross-positio-ning system were analyzed． The simulation results show that the FDA can well deduct the effect of the interferome-ter in the far field condition by reasonably selecting the carrier frequency，the number of array elements and the
frequency offset． While other parameters are reasonably selected，the higher the direction finding accuracy is，the smaller the positioning blur area is，and the position of the positioning virtual jammer always falls outside the 50% CEP radius．


Cooperative Network Coding in VANETs Based on Distributed TDMA



OU Mang WANG Jiwen

2020, 48(1):  104-113.  doi:10.12141/j.issn.1000-565X.190150

Asbtract ( 268 )   HTML ( 6)   PDF (1088KB) ( 625 )  

Related Articles | Metrics

A cooperative network coding communication method based on distributed TDMA ( time division multi-ple access) was proposed for the VANETs uplink application scenario． The cooperative node uses the MIMO_NC technology to encode its own service data and packets that failed to be transmitted by other nodes，so as to coopera-tively retransmit the failed data packets while transmitting their own service data． The proposed method utilizes the intrinsic mechanism of distributed TDMA to confirm the data transmission result，and uses the message carrying mechanism to exchange relevant control information to avoid sending special control packets． The theoretical and experimental results show that the proposed method can significantly improve the packet reception probability and effectively reduce the packet transmission delay and packet dropping rate．


Iterative Learning Control for Linear Differential Repetitive Processes with Regional Pole Constraints#br#



WANG Lei, YANG Huizhong, TAO Hongfeng, et al

2020, 48(1):  114-122.  doi:10.12141/j.issn.1000-565X.190195

Asbtract ( 276 )   HTML ( 3)   PDF (899KB) ( 164 )  

Related Articles | Metrics

Existing studies on iterative learning control focus on the robust convergence of the iterative learning control system along the trial direction，and pay little attention to the performance along the time direction，which fails to show the dynamic characteristics of the control system． Iterative learning control for linear differential repet-itive processes with regional pole constraints was studied． Firstly，combining with the two-dimensional (2D) sys-tem theory，the 2D continuous Roesser model was built under the time domain． Then，based on the Kalman-Yakubovich-Popov lemma，the tracking performance of iterative learning control systems and regional pole con-straints problems were analyzed． Moreover，the sufficient conditions of existence for the controller were derived in terms of linear matrix inequalities，which can guarantee the system performance along the trial direction and the time direction． Furthermore，the results are extended to structural uncertainty model as well． Finally，the simula-tions for a typical actuator of tracking servo system prove that the design is effective and feasible．


Research on Swing Amplitude Detection of Automobile Wiper with Two Granularity Optical Flow Manifold Learning#br#



ZHENG Sifan, WANG Weixing, HE Zhanhua, et al

2020, 48(1):  123-132.  doi:10.12141/j.issn.1000-565X.190175

Asbtract ( 289 )   HTML ( 7)   PDF (5140KB) ( 499 )  

Related Articles | Metrics

In the vehicle security inspection system design based on machine vision，a subspace clustering algo-rithm via two granularity optical flow manifold learning was proposed to detect automobile wiper swing angle ampli-tude in order to avoid the defect that the optical flow trajectory is too sparse caused by the complex background of glass． Firstly，the full-length wiper LDOF variational optical flow trajectory was used as coarse-granularity optical flow to perform sparse subspace clustering to obtain reliable seed sample． Then the temporal-spatial similarity to-pography graph of dense and fine granularity optical flow and coarse-granularity optical flow was constructed，and the harmonic function was used to relax the neighbor trajectory label of the seed's to the Gaussian random field for semi-supervised label diffusion to obtain a dense wiper motion region for further RANSAC line fitting and swing an-gle calculation． Finaly，the algorithm module was encapsulated by ocx plugin and embeded into the vehicle track-ing module in the form of callback function for synchronization． Security check video of 153 trains of 4 types of ve-
hicles under 6 different illuminances was collected． The videos was used to analyze and compare the accuracy of fitting and swing angle of two granularity optical flow manifold learning algorithms． The experiment results show that the accuracy of fitting and swing angle of the new algorithm can reach more than 85%，showing a broad appli-cation and promotion prospect．


Channel Estimation Method Based on Multi-way Array in CCFD Systems



HAN Xi, ZHOU Yingchun, ZHAO Xinyuan, et al

2020, 48(1):  133-138,146.  doi:10.12141/j.issn.1000-565X.190138

Asbtract ( 443 )   HTML ( 9)   PDF (563KB) ( 94 )  

Related Articles | Metrics

A channel estimation method based on multi-way array is proposed based on a two-way co-time co-fre-quency full-duplex (CCFD) relaying system． The method consists of three parts: self-interference cancellation，multi-way matrix modeling and channel estimation． At the sides of the two users，the proposed method simultane-ously transmits the designed pilot sequences to the relays，and the relays performs self-interference cancellation and then amplify and forward the received signals to two users． The proposed channel estimation method does not re-quire channel estimation and iteration at the relays，thus reduce the burden of the relays． Simulation results show that compared with the channel estimation methods of the existing Least Squares (LS)，Two-Stage Training (TST) and non-iterative P_KRF，the proposed method has higher channel estimation accuracy，and it can oper-ate under more flexible antenna configurations，which is more stable and more practical in the future application．


Lip Motion and Voice Consistency Recognition based on Specific Vowel Pronunciation Events Analysis



ZHU Zhengyu, QIU Huayu, YANG Chunling, et al

2020, 48(1):  139-146.  doi:10.12141/j.issn.1000-565X.190287

Asbtract ( 249 )   HTML ( 2)   PDF (1140KB) ( 205 )  

Related Articles | Metrics

The traditional lip motion and voice consistency recognition method is to analyze the whole sentence without filtering the content，which is complicate in computation and its results are vulnerable to weak related segments such as mute． The vowels which with significant lip shape changes were researched in depth． By analyzing the audio and lip motion correlation of each vowel category clustered by lip sequence features，a more representative specific phonological pronunciation unit was selected as the analysis object． Combined with audio-visual delay analysis，a consistent recognition method based on specific vowel pronunciation events analysis was proposed．Firstly，the selected unit was segmented and identified． Then the correlation degree of each specific vowel event was obtained，and the delay distribution of each specific vowel occurrence position was statistically scored． Finally，a consistency judgment was made by combining the vowel pronunciation event audio-visual correlation score with the position delay analysis score． Compared with other methods through experiments，results show that the proposed method is superior in recognition performance and reduces the amount of computation．