Table of Content

30 June 2024, Volume 30 Issue 3

Previous Issue   

Ecofriendly phosphorus-tin composite synergistic flame retardant: preparation and application in polyvinyl chloride

CHEN Liquan, HONG Jianting, NING Lixin, PENG Yuqing, LI Aijun, MA Dianpu, WANG Jinhe, LIU Liqi

2024, 30(3):  373-389.  doi:10.12066/j.issn.1007-2861.2404

Asbtract ( 270 )   PDF (10684KB) ( 160 )  

Related Articles | Metrics

This article presents a synthetic method for tin phytate-coated zinc hydrox-ystannate (ZHS@Sn-Phyt) particles with different ratios, which were incorporated into soft polyvinyl chloride (PVC) as a composite synergistic flame retardant. The synthe-sized ZHS@Sn-Phyt particles were characterized by X-ray diffraction, scanning electron microscopy, Fourier transform infrared spectroscopy, X-ray photoelectron spectroscopy, and thermogravimetric analysis. The results revealed that a stable coating structure was formed in the form of Sn—O—P—O—Zn ionic bonds at the interface between the zinc hydroxystannate (ZHS) surface and the tin phytate (Sn-Phyt) coating layer. The PVC composites were evaluated through limiting oxygen index (LOI), cone calorimetry, and mechanical studies. When the mass ratio of ZHS@Sn-Phyt in the flame retardant was 1∶9, the best flame retardant properties of PVC were achieved, indicating that combining phosphorus and tin results in a better synergistic flame retardant effect. The LOI of the PVC/ZHS@Sn-Phyt composite was 1.6% higher than that of the PVC composite with only ZHS (without the Sn-Phyt coating). Moreover, the peak heat release rate and total heat release decreased by 18.9% and 31.7%, respectively. In addition, the carbon residue rate increased by 6.1% and the total smoke release of PVC/ZHS@Sn-Phyt and PVC/ZHS were similar. Both the tensile strength and the elongation at break of PVC/ZHS@Sn-Phyt were also better than those of PVC/ZHS at 26.79 MPa and 296.5%, respectively. Thus, this study provides a method for preparing efficient and green flame retardants for PVC. The phosphorus/tin-based synergistic flame retardant prepared by our method has the advan-tageous characteristics of ecofriendliness and simplicity of the preparation process.



Mechanisms of three damping models and their influence on dynamic characteristics of drill strings

DI Qinfeng, CHEN Gong, HU Feifei, WANG Wenchang, CHEN Feng

2024, 30(3):  390-400.  doi:10.12066/j.issn.1007-2861.2550

Asbtract ( 251 )   PDF (609KB) ( 81 )  

Related Articles | Metrics

The Rayleigh damping model is widely used in simulating the dynamic charac-teristics of drill strings to reflect the effect of damping (including drilling fluid damping). However, the basis for choosing this model has not been sufficiently investigated, and its damping coefficient is mainly determined by experience. In this study, the dynamic fi-nite element model of a drill string was established by first considering the nonlinearity and damping effect of the drilling fluid to simulate the dynamic characteristics of the drill string in a borehole. Subsequently, the mechanisms and internal relations of the typical Rayleigh damping, Caughey damping, and Clough damping models were investigated. Finally, the effects of these three models on the dynamic characteristics of the drill string were investigated using a practical example, and the lateral displacement, lateral acceleration, whirl trajectory, and whirl velocity of the drill string under different damping models were compared. The results show that the damping value of the Caughey damping model with l within the range of −2 ∼ 1 is negligible to effectively provide viscous damping, and the damping value is extremely large when l assumes other values. This implies that the Caughey damping model with four terms is not suitable for the dynamic simulation of drill strings. The Rayleigh and Clough damping models can effectively reflect the damping effect of drilling fluid, and the dynamic characteristics of the drill string obtained are consistent with the actual vibration. By contrast, the Rayleigh damping model is simple in form and easy to calculate; therefore, it is suitable for the dynamic simulation of drill strings in drilling engineering.



Analysis of the successive excavation disturbance of three-line shield tunneling

SHI Xiaoyi, ZHANG Mengxi, KE Ningjing, LU Kangming

2024, 30(3):  401-413.  doi:10.12066/j.issn.1007-2861.2465

Asbtract ( 163 )   PDF (1935KB) ( 61 )  

Related Articles | Metrics

In a three-line parallel shield tunnel, the construction of a second shield tunnel may affect existing advanced tunnels. An analytical solution for the additional horizontal displacement of the leading tunnel after two parallel side passes of the shield was obtained by establishing a two-stage analysis model and using the Mindlin solution. The effectiveness of the method was verified using measurement data from the Shanghai X-ray shield tunnel. A parametric analysis was performed to investigate the effects of the horizontal distance between the tunnels, longitudinal stiffness of the tunnel structure, shear modulus of formation, and ground loss on the tunnel responses. The results showed that the additional displacement peak generated by the first tunnel under the parallel side crossing of the second tunnel was inversely proportional to the horizontal distance of the parallel tunnel, longitudinal stiffness of the tunnel structure, shear modulus of the formation, and formation loss rate in direct proportion. The proposed method can provide a theoretical basis for predicting the influences of two parallel side passes of a shield on an existing advanced tunnel in relevant engineering practice.



ESG spillover effect of supply chain based on the perspective of upstream and downstream resource sharing

LI Yuanqin, LI Tong, LI Ting

2024, 30(3):  414-434.  doi:10.12066/j.issn.1007-2861.2544

Asbtract ( 342 )   PDF (433KB) ( 1330 )  

Related Articles | Metrics

A-share listed companies from 2010 to 2022 as a sample was used to investigate the spillover effect of ESG (environment、social and governance) on supply chain in the context of supply chain integration. The results showed that the ESG levels of suppliers and customers could improve the ESG levels of enterprises. The ESG spillover effect mainly occured through three channels: management knowledge resources, capital resources and technology resources sharing. For non-state-owned enterprises and regional enterprises with a higher degree of marketization, the spillover effect of ESG was more obvious. This study has important reference value for implementing strategic decision to maintain supply chain security in China, and is significant for promoting the high-quality development of listed companies.



UAV signal recognition method based on re-classification and separation of image transmission signal and flight control signal

WANG Anping, LV Zhenbin, YI Zixuan, SHEN Huaming, HUANG Jiapeng, LU Wenbin

2024, 30(3):  435-450.  doi:10.12066/j.issn.1007-2861.2461

Asbtract ( 313 )   PDF (13885KB) ( 73 )  

Related Articles | Metrics

The rapid development of the unmanned aerial vehicle (UAV) industry has introduced security risks in low-altitude airspaces. To effectively control UAVs, a radio detection system that can identify UAV signals should be developed. To identify similar UAVs, this study proposes a UAV signal recognition method based on reclassification and separation of image transmission signals (ITSs) and flight control signals (FCSs). The pro-posed method extracts its time-domain parameters using the cyclic characteristics of the ITS and applies a classification decision tree to initially classify and identify the UAVs. Then, by separating the ITS and FCS, their time-frequency characteristic parameters are extracted, and finally, secondary classification is performed. Experimental results show



Non-stationary wind velocity simulation using deep reinforcement learning-based regulation and control

CAO Liyuan, ZHANG Zhenyu, LI Chunxiang

2024, 30(3):  451-465.  doi:10.12066/j.issn.1007-2861.2569

Asbtract ( 163 )   PDF (8754KB) ( 75 )  

Related Articles | Metrics

A novel hybrid simulation method for a deep deterministic policy gradient (DDPG) algorithm and generalized S-transform (GST), referred to as DDPG-GST, is pro-posed. In the DDPG-GST method, empirical mode decomposition is first used to decompose the original data into nonstationary fluctuating wind speed components and trend components. The GST is then used to extract the time–frequency characteristics of the nonstationary fluctuating wind speed components, followed by the construction of the GST time–frequency power spectrum matrix. Subsequently, Cholesky decomposition is applied to generate simulated nonstationary fluctuating wind speeds. These simulated speeds are input into the DDPG network for regulation and control to optimize the simulation pro-cess. Finally, the simulated total wind speeds are obtained by superposing the simulated nonstationary fluctuating wind speeds with the trend components. The results show that DDPG-GST retains the energy characteristics of nonstationary fluctuating wind speeds more accurately in the time domain compared to the GST simulation method. Additionally,the energy distributions, derived from the GST coefficient amplitudes by the DDPG-GST method in the time-frequency domain, align more closely with the targets. The average power spectrum of the DDPG-GST method is closer to the target. Therefore, the non-stationary wind speed simulation based on deep reinforcement learning is a high-precision, data-driven simulation method.



Lightweight image compression algorithm based on deep learning

FAN Shenwei, LI Guoping, WANG Guozhong

2024, 30(3):  466-475.  doi:10.12066/j.issn.1007-2861.2467

Asbtract ( 319 )   PDF (2083KB) ( 311 )  

Related Articles | Metrics

The transformation modules of image compression algorithms based on deep learning involves complex architectures and large quantities of computation. To speed up the encoding and decoding process, a method was proposed to reduce the number of parameters and multiply-accumulation operations (MACs) of the original network with knowledge distillation while maintaining the image quality as much as possible. The original and the lightweight networks were trained simultaneously, and the lightweight network performance was improved by receiving feature information from the original network. When designing the lightweight network, group convolution was introduced to retain more feature information and reduce the number of parameters and MACs of the network as much as possible, while the number of channels was reduced. Experiments on the test datasets Kodak and DIV2K showed that, compared with the original network, the lightweight network after knowledge distillation still maintained good image quality while the amount of parameters and MACs was approximately one-sixteenth that of the original network.



Anti-noise speech recognition system based on generative adversarial network data enhancement

FENG Tianyu, ZHU Yonghua

2024, 30(3):  476-490.  doi:10.12066/j.issn.1007-2861.2449

Asbtract ( 301 )   PDF (327KB) ( 218 )  

Related Articles | Metrics

Research on speech recognition is always challenged by the limitations of the dataset. Data enhancement can improve the scale and diversity of training data, thereby improving the accuracy of speech recognition. In this paper, a speech data generation method based on generative adversarial network (GAN) is proposed for improving speech recognition in noisy environments. First, the basic GAN structure is used to generate speech samples frame by frame at the spectral feature level. Considering the lack of real labels for training, an unsupervised learning framework is proposed for acoustic modeling using non-transcribed data, whereby the conditional GAN structure is used to explore two conditions: the acoustic state of each speech frame and original clean speech corresponding to the speech in the dataset. GANs that incorporate conditional information can directly provide real labels for acoustic modeling. The present method was evaluated on the noisy Aurora-4 and AMI conference transcription tasks. Experimental results show that the new method can significantly improve the performance under various noise conditions (additive noise, channel distortion, and reverberation). The enhanced data generated by GAN reduced the word error rate (WER) by 6%∼14% on the advanced very deep convolutional neural network (VDCNN) acoustic model. 



Virtual try-on network based on pose guidance

HUANG Dongjin, LI Xiaomin, LIU Jinhua, LI Zhenyan

2024, 30(3):  491-502.  doi:10.12066/j.issn.1007-2861.2466

Asbtract ( 202 )   PDF (8552KB) ( 24 )  

Related Articles | Metrics

Existing pose guidance-based virtual try-on methods suffer from excessive deformation of clothing texture and occlusion of generated images. To address these issues, this paper proposes a pose-guided virtual try-on network (PG-VTON), an improved virtual try-on network based on the Downto network. First, a U-Net-based network is used to transform the pose of a figure image and generate the figure parsing image under the target pose. An information-enhancement module is introduced to improve the accuracy of the parsing image. Then, a thin plate spline (TPS) is used to transform the target cloth into a shape that fits the body of the target figure, and grid warping regularization is introduced to preserve the texture and detail features of the target cloth. The final virtual try-on image is generated by combining the parsing image with a warped cloth. Experimental results indicate that the proposed method improves the average structural similarity (SSIM) of the virtual try-on image by 2.83% and inception score (IS) by 6.74%compared with the Downto network. Further, as compared to other virtual try-on methods, the proposed method reduces false occlusion in the process and generates clearer and more realistic results.



Reconfigurable design of state secret algorithm SM4 and ZUC based on FPGA

LI Yanran, YAN Limin

2024, 30(3):  503-511.  doi:10.12066/j.issn.1007-2861.2434

Asbtract ( 196 )   PDF (2489KB) ( 64 )  

Related Articles | Metrics

This paper proposes a hardware implementation scheme of SM4 and ZUC-256 ciphers. This scheme first provided hardware language description for the two ciphers, and then two reconfigurable units—reconfigurable register and reconfigurable S-box unit were constructed. Accordingly, based on the realization of two password algorithms, resource consumption was effectively reduced and resource utilization was improved. Hardware veri-fication was carried out using a field programmable gate array (FPGA). The results showed that the hardware overhead in the password reconfiguration scheme was significantly re-duced.



First-principles calculation of radiation resistance and mechanical properties of tungsten/graphene/tungsten and tungsten/graphane/tungsten as plasma facing materials

YAN Qizhen, ZHANG Zhaochun, GUO Haibo, WANG Yang

2024, 30(3):  512-521.  doi:10.12066/j.issn.1007-2861.2432

Asbtract ( 163 )   PDF (10436KB) ( 75 )  

Related Articles | Metrics

Tungsten/graphene/tungsten and tungsten/graphane/tungsten nanolayered composites were developed and applied as the plasma facing materials of nuclear fusion devices. The formation energies of the helium interstitial atom and tungsten vacancy and the mechanical properties of the two nanolayered composites were investigated by first-principles calculation. The results indicate that the tungsten/graphene/tungsten and tungsten/graphane/tungsten nanolayered composites exhibit resistance to radiation dam-age to some extent. The tungsten/graphene and tungsten/graphane interfaces serve as sinks for both the helium interstitial atom and tungsten vacancy and as sites for the self-healing performance of tungsten self-interstitial and vacancy. The results based on the Fukui function reveal that these features of the tungsten/graphene and tungsten/graphane interfaces are associated with the change in the activity of tungsten atoms. An interesting finding of this study is that the ductility of tungsten-based materials can be enhanced and the anisotropy factor reduced by inserting monolayer graphene into the tungsten crystal.



Motion modeling and control of a 2-DOF soft manipulator based on a recurrent neural network

DING Wei, ZHENG Yun, ZHONG Songyi, YANG Yang

2024, 30(3):  522-531.  doi:10.12066/j.issn.1007-2861.2479

Asbtract ( 177 )   PDF (3770KB) ( 212 )  

Related Articles | Metrics

To address the difficulty of modeling and control of existing soft manipulators due to their small material stiffness and unstable modulus, this study proposes a method based on a recurrent neural network (RNN) for the motion modeling of a two-degree-of-freedom (2-DOF) soft manipulator with control. A motion-capture instrument was used to collect the position coordinates under different pressures and loads, and the coordinates were imported into a gated recurrent unit (GRU) neural network model for training. The accuracy of the test set reached 98.87% when the hyperparameters were adjusted to the optimal network structure. Accordingly, a mapping function for the pressure and load at the end position was constructed. Experimental results showed that the proposed method could improve the control accuracy of the manipulator by approximately 6∼8 mm and significantly reduced the difficulty of control and modeling of a soft robot.



Flow characteristics of elliptical column under subcritical Reynolds numbers

WU Yongxi, ZHANG Xiangyu, WU Hao, ZHANG Dan, MAO Fangsai

2024, 30(3):  532-544.  doi:10.12066/j.issn.1007-2861.2416

Asbtract ( 200 )   PDF (4707KB) ( 336 )  

Related Articles | Metrics

This study explores the characteristics of cylindrical winding flow at varying axis ratios (AR) within subcritical Reynolds numbers. We examine the flow’s lift coefficient, drag coefficient, and St at critical Reynolds numbers of Re=3.9×10³, 5.0×10³, 6.0×10³, 8.0×10³, and 10.0×10³, with AR ranging from 0.1 to 1.6. We analyze the causes of flow characteristics by examining the vorticity cloud map and pressure coefficient distribution. The findings indicate that at subcritical Reynolds numbers, maximum lift coefficient CL.m varies with different Reynolds numbers, displaying a trend of initial increase followed by a decrease as AR increases, peaking at AR = 0.7. The trend in St numbers across various Reynolds numbers is consistently upward. The mean drag coefficient exhibits two distinct trends at a critical point when AR = 0.7. These results offer valuable insights for the optimal design of devices utilizing piezoelectric configurations.



An optimization method based on support vector machine for Ramachandran plot in protein structures annotation

WANG Bo, SU Tianhao, XU Yanting, GAO Heng, GUO Cong, LI Yongle, WU Wei

2024, 30(3):  545-558.  doi:10.12066/j.issn.1007-2861.2462

Asbtract ( 96 )   PDF (1294KB) ( 210 )  

Related Articles | Metrics

The Ramachandran plot is among the most central concepts for validating the conformation of protein structures, and accordingly plays an important role in structural biology. However, the favored regions defined when using the traditional Ramachandran plot are too wide and contain inaccurate structures. To address these deficiencies, a method based on support vector machine (SVM) and Bayesian optimization (SVM-Rama) for optimization and subdivision of the definition of favored regions for the Ramachandran plot is proposed. Aims in this study are to enhance the accuracy of the favored regions for the specific secondary structure species of proteins, and subsequently to validate and annotate protein secondary structures simply and accurately. The results reveal that the optimized plot has a high accuracy comparable to the best performance of traditional methods in secondary structure annotation, while facilitating analysis at lower training and computational costs than these traditional methods.



Deep accumulation of magnetic microspheres under the combined action of magnetic field and ultrasonic wave

PEI Ning, HAN Menglong, GONG Yongyong, DONG Yaocheng, SHEN Yunli

2024, 30(3):  559-564.  doi:10.12066/j.issn.1007-2861.2431

Asbtract ( 179 )   PDF (3888KB) ( 56 )  

Related Articles | Metrics

In magnetic drug targeting (MDT), a large quantity of magnetic drug particles is absorbed near the magnetic pole, with only a small number of those adsorbed far away from the magnetic pole, where the magnetic field produced by the general electromagnet is attenuated, hindering MDT applications in deep tissue. In this paper, we propose a new noninvasive approach, in which the magnetic field is combined with ultrasound to achieve three-dimensional deep accumulation of magnetic particles. First, the structure design of the experimental equipment for deep aggregation of magnetic particles is presented. Second, the range of ultrasonic amplitudes that can achieve deep aggregation is calculated using the COMSOL simulation software. Finally, an experiment is performed based on simulations and experimental results show that the quantity of magnetic microspheres absorbed in areas far from the magnetic poles is more than those in areas close to the magnetic pole, thereby demonstrating that deep aggregation of magnetic particles was achieved.