Table of Content

25 May 2025, Volume 53 Issue 5

Previous Issue   

Mechanical Engineering

Hybrid Toolpath Planning and Feedrate Optimization for Complex Pockets Machining Under Constant Loads

WANG Qinghui, WANG Jingqiang, DING Xuesong, et al

2025, 53(5):  1-10.  doi:10.12141/j.issn.1000-565X.240467

Asbtract ( 83 )  

Related Articles | Metrics

The CNC machining of molds and various 3D parts involves numerous pocket features, and the design of machining toolpaths directly affects machining quality and efficiency. With advancements in high-speed milling technology, CNC machines provide the hardware foundation for improving pocket machining efficiency, but they also place higher demands on CAM toolpath design. Traditional CAM toolpaths tend to cause abrupt changes in cutting load when dealing with areas such as pocket corners, slots, and intersections of circular paths. This load instability limits the improvement of feed rate and cutting depth, negatively impacting both machining efficiency and quality. To address these issues, this paper proposes a combined toolpath design and optimization method aimed at achieving constant load machining for pockets. The method is based on a multi-level block structure, first calculating the material removal rate (MRR) and then dividing the machining areas into stable, semi-stable, and load fluctuation regions. Different strategies are applied to each region, including circular toolpaths, feed speed optimization, and variable-radius trochoidal paths, to ensure smooth load control throughout the machining process.By applying trochoidal paths in areas prone to load fluctuations, the method reduces sudden load variations and ensures the stability of the machining process. Experimental results show that the proposed design and optimization method is suitable for generating CAM toolpaths for various complex pockets, ensuring load stability and improving machining quality.

Design and Study of Quick Coupling Device Used for Emergency Rescue Robot Attachments

LI Wei, LIU Jiachen, ZHANG Weiyuan, et al

2025, 53(5):  11-19.  doi:10.12141/j.issn.1000-565X.240408

Asbtract ( 56 )  

Related Articles | Metrics

The existing emergency rescue equipment has the problem of single function and low flexibility, which is difficult to meet the needs of emergency operations under earthquake, geology and other disaster conditions. A kind of electromechanical hydraulic quick coupling device is studied, which can realize fast change of attachments and free adjustment of position and pose. The device can be quickly integrated into emergency rescue equipment to complete the high-mobility multi-function rescue task. The extreme load characteristics, stress and strain situations during operation of the device are analyzed. The weak part and load spectrum are determined. The reliability theoretical models considering cyclic damage strength degradation under deterministic and random periodic stress are deduced and established. The mapping relationship between the reliability and failure rate of the rotating mechanism is determined. Based on linear elastic fracture mechanics, the crack propagation of the weak part is analyzed. The fatigue life of the device is determined by the local stress strain method to ensure that the device can meet the application requirements. The developed device is integrated into the walking rescue robot. The test results show that it can realize the rapid switching of various attachments such as bucket and gripper, with a switching time of less than 15s. The increased ±40° yaw and ±360° rotation degrees of freedom can meet the needs of flexible operation. The research results can provide significant theoretical reference for the design of similar devices, which is of great significance for improving China's disaster rescue and emergency response capabilities.




Machining feature recognition method of B-Rep model based on graph neural network

HU GuangHua, DAI ZhiGang, WANG QingHui

2025, 53(5):  20-31.  doi:10.12141/j.issn.1000-565X.240329

Asbtract ( 135 )  

Related Articles | Metrics

Automatic feature recognition (AFR) is one of the key technologies of intelligent manufacturing. Traditional rule-based recognition algorithms have poor scalability, while methods based on deep convolutional networks use discrete models as input, have low accuracy, and the recognition results are difficult to accurately map back to the original CAD model, causing inconvenience in application. In view of the above shortcomings, we proposed a feature recognition method based on graph neural network, which can directly analyze B-Rep models. Our method extracts effective characteristic information and geometric information from the B-Rep structures to form a feature descriptor, and then establishes an adjacency graph consist of high-level semantic information based on the topological structure of the CAD model. Taking the adjacency graph as input, an efficient graph neural network model is constructed. By introducing a differentiable generalized message aggregation function and a residual connection mechanism, the model has stronger information aggregation performance and multi-level feature capture capabilities. What is more, the message normalization strategy is used to ensure the stability of the training process and to accelerate the convergence of the model. After training, the network can directly classify and annotate all faces in the B-Rep model, thereby realizing feature recognition. Experimental results based on the public dataset MFCAD++ demonstrates that the proposed method achieved an accuracy of 99.53% and an average intersection-over-union ratio of 99.15%, which outperforms other similar studies. Further evaluations using more complex testing cases and typical CAD cases from real engineering applications show that the proposed method has better generalization ability and adaptability.

Mechanism of Saw-Tooth Shaped Chip Formation in Cast And Forged Superalloy GH4198

JIN Qichao, LI Jun, WAN Liangliang, et al

2025, 53(5):  32-44.  doi:10.12141/j.issn.1000-565X.240390

Asbtract ( 31 )  

Related Articles | Metrics

The purpose of this study is to analyze the chip formation during the turning of process of a new high-temperature alloy GH4198, predict chip morphology using theoretical models, and elucidate the formation mechanism of saw tooth-shaped chips. The methodology involves conducting right-angle cutting experiments to record cutting forces, collecting chip samples, measuring geometric shapes, and calculating the shear angle. The influence of the tool bluntness radius on chip formation was analyzed, and the geometric shape of serrated chips was predicted based on utilizing slip line field theory. Orthogonal experiments were designed to analyze the effects of cutting speed and feed rate on the geometric shape of serrated chips. A three-stage physical model of serrated chip formation, considering the tool bluntness, was proposed based on adiabatic shear theory. The Johnson-Cook constitutive model and fracture criterion were chosen to establish a two-dimensional orthogonal cutting finite element model with thermal coupled, and its validity was confirmed by comparing experimental and simulated cutting force data. The mechanism of saw tooth chip formation was analyzed based on stress, equivalent plastic strain, and temperature changes during chip formation. Two-dimensional orthogonal cutting finite element models with varying tool bluntness radii were established. By examining the impact of tool bluntness radius on parameters such as stress and strain during chip formation, the study explored its influence on the shaping of serrated chips. The research found that as cutting speed and feed rate increased, the shear angle also increased, while chip thickness decreased with higher cutting speeds. At cutting speeds ranging from 10 to 30m/min, the relative error in predicted chip thickness ranged from 4.20% to 24.73%, and the relative error in simulated cutting forces ranged from 4.19% to 9.14%. The maximum compression ratios of chip thickness were 3.19, 2.78, and 2.26, with serration levels of 0.20, 0.36, and 0.58, respectively. At a cutting speed of 30m/min, chips exhibited significant cracks and an overall tilt in the saw tooth profile. With a feed rate of 0.05 to 0.15mm/r, the relative errors in predicted chip thickness were between 5.07% and 17.66%, and the relative errors in simulated cutting forces ranged from 6.42% to 14.23%. The maximum compression ratios of chip thickness were 2.82, 2.78, and 2.61, with serration levels of 0.12, 0.36, and 0.42, respectively. An increase in the tool bluntness radius led to a higher rate of material accumulation near the tool tip per unit time, exacerbating plastic deformation in the primary deformation zone. This enhanced the plowing and squeezing actions of the tool during cutting, causing increases in cutting force, stress, equivalent plastic strain, and temperature, thereby resulting in a greater degree of serration in the chips. The study concluded that the slip line field model could effectively predict the variation in chip thickness with changes in cutting parameters. As cutting parameters increased, the compression ratio of chip thickness showed a decreasing trend, while the degree of serration increased. The finite element model was utilized to analyze the patterns of equivalent plastic strain and temperature changes during chip formation, as well as the impact of the tool fillet radius on chip formation. The validity of the theoretical model of chip formation was confirmed, providing a theoretical basis and practical guidance for improving the integrity of machined surfaces.

Analysis of Extrusion Law of Large and Small Aluminum Profiles with Hollow Thin Wall for Track#br#

LIU Guoyong, GAO Shize, ZHU Dongmei

2025, 53(5):  45-55.  doi:10.12141/j.issn.1000-565X.240397

Asbtract ( 33 )  

Related Articles | Metrics

In order to explore the extrusion law of large and small hollow thin-wall aluminum profiles for track, hyperxtrude simulation software is used to numerically simulate the extrusion process of hollow thin-wall aluminum profiles for track, explore the influence of mold structure and process parameters, and compare the forming rules of two similar shapes of large and small profiles. In terms of mold structure, the modification of welding chamber and drainage groove has the most obvious influence on large and small profiles. Through simulation analysis, it is found that changes in the welding chamber significantly reduce the maximum deformation of the large-profile material, with a reduction of 25.34%, while the for the small-profile material is 42.82%. At the same time, changes in the structure of the drainage groove show different impact trends; after altering the drainage groove, the maximum deformation reduction for the large-profile is 40.88%, while that of small-profile is 24.72%. The results show that because the drainage groove of small profile is shorter, and the modification of the drainage groove of large profile is more complicated, the change of drainage groove has a more significant impact on large profile. In terms of process parameters, the changes of metal deformation, metal flow rate differences, and the SDV values of the exit section of the profile under different conditions were analyzed. The results shows that the extrusion speed and die temperature have a more significant impact on the large-profile, while the billet diameter has a more pronounced effect on the small-profile. This finding provides important theoretical support for optimizing the extrusion process of aluminum profiles.

Computer Science & Technology

A Method for Software Vulnerability Detection via Path Representations and Pretrained Model

LU Lu, WAN Tong

2025, 53(5):  56-65.  doi:10.12141/j.issn.1000-565X.240131

Asbtract ( 60 )  

Related Articles | Metrics

Software vulnerabilities represent critical vulnerabilities that can compromise system security and are susceptible to exploitation by attackers for unauthorized control. Contemporary deep learning-based vulnerability detection approaches largely suffer from limitations due to their reliance on single code representations, failing to fully capture the complementary nature of code semantics and structural information. This research introduces an innovative method for software vulnerability detection, termed VDPPM (Software Vulnerability Detection via Path Representations and Pretrained Model), which addresses this issue. The proposed framework integrates path representations extracted from Abstract Syntax Tree (AST), Control Flow Graph (CFG), and Program Dependency Graphs (PDG), thereby offering a more comprehensive view of code characteristics. The VDPPM framework employs SimCodeBERT, a model refined through contrastive learning framework SimCSE, enhancing its ability to interpret code semantics. In the experimental phase, we initially construct a corpus using path representations and train a Doc2vec model to generate general-purpose embedding models, converting sequence of paths into vector representations. Subsequently, a pretrained CodeBERT model is integrated, which, after training under the contrastive learning framework, gains increased precision in capturing deep semantic features within the code. Ultimately, the fusion of vector representations generated by both Doc2vec and the enhanced SimCodeBERT enables the effective execution of vulnerability detection. Empirical studies demonstrate that across multiple publicly available benchmark datasets for vulnerability detection tasks, the VDPPM framework outperforms mainstream methods, showing significant improvements in several performance metrics. This convincingly validates the effectiveness and superiority of the proposed methodology.

Research on Carbon Emission Forecasting in Transportation Industry Based on SD-ISSA-DALSTM

WANG Qingrong, WANG Junjie, ZHU Changfeng, et al

2025, 53(5):  66-81.  doi:10.12141/j.issn.1000-565X.240356

Asbtract ( 144 )  

Related Articles | Metrics

Aiming at the problem of high volatility and stochasticity and low prediction accuracy in the carbon emission data series of transportation industry, a transportation carbon emission prediction model combining the quadratic decomposition （SD）, dual attention mechanism （DA）, improved sparrow search algorithm （ISSA） and LSTM network is proposed. First, CEEMDAN is introduced to decompose the transportation carbon emission data series into modal components with different frequencies, and then the sample entropy （SE） is used to quantify the complexity of each component, and the secondary decomposition （SD） is performed on the component with the highest entropy value by using VMD, which further weakens the volatility and nonlinearity of the transportation carbon emission data series; second, in order to tap the correlation between the transportation carbon emission and its influencing factors, a feature-attention mechanism is added to the inputs of the LSTM Secondly, in order to explore the correlation between transportation carbon emissions and its influencing factors, a feature attention mechanism is added to the input side of the model to highlight the key input features; meanwhile, a temporal attention mechanism is added to the output side to extract the key historical moments; finally, the SSA algorithm is improved by combining the Circle chaotic mapping, dynamic inertia weight factor and mixed variance operator strategies, and the ISSA-DALSTM models are established for each component separately, and then reconstructed for the predicted values of each component. The carbon emission data of China's transportation industry from 1990 to 2019 are measured to validate the model, and the results show that the RMSE, MAE, and MAPE of the proposed model are 5.3088, 3.5661, and 0.4439, respectively, which are better than those of other comparative models, and the validity of the proposed model is verified.

A Spatiotemporal Heterogeneous Two-Stage Fusion Network for Traffic Flow Forecasting

HOU Yue, YIN Jie, ZHANG Zhihao, et al

2025, 53(5):  82-93.  doi:10.12141/j.issn.1000-565X.240480

Asbtract ( 97 )  

Related Articles | Metrics

In response to the existing traffic flow prediction studies that fail to fully integrate complex spatiotemporal correlations and heterogeneities, this paper designs a traffic flow prediction network based on grid data—the Spatiotemporal Heterogeneous Two-Stage Fusion Neural Network (ST_HTFNN). This network employs a phased, hierarchical spatiotemporal feature extraction architecture, adopting a new model where the static and dynamic feature extraction stages are serialized. In the static feature extraction stage, a novel Mamba-Like Linear Attention (MLLA) block is introduced as a static heterogeneous fusion unit to achieve spatial correlation and heterogeneity fusion mining. In the dynamic feature extraction stage, a simple and efficient dynamic heterogeneous fusion unit is designed, combining dilated convolution with gating mechanisms to adaptively fuse and capture global and local spatiotemporal correlations and heterogeneities. Furthermore, to address the smoothing of road features during the deep convolution process at the road-level traffic flow characteristics, a road feature enhancement module is designed to reconstruct and enhance road information. Finally, an external disturbance feature fusion module is designed to integrate the impact of external disturbance features on traffic flow prediction results. Experimental results on three real-world traffic datasets—BikeNYC, TaxiCQ, and TaxiBJ—demonstrate that the ST_HTFNN model outperforms existing benchmark methods, with an average improvement of 6.13%, 2.06%, and 5.23% in the prediction accuracy evaluation metric MAE.

A Liver Tumor Segmentation Method Based on Multi-Scale Feature Fusion and Reconstruction Convolution

MA Jinlin, JIU Zhiqing, MA Ziping, et al

2025, 53(5):  94-108.  doi:10.12141/j.issn.1000-565X.240439

Asbtract ( 71 )  

Related Articles | Metrics

Aiming at the problem of insufficient expression ability of liver tumor features and limited global contextual information transmission, an improved UNet liver tumor segmentation method is proposed. Firstly, a low rank reconstruction convolution is designed to optimize the large number of parameter problems caused by traditional convolution operations, and use it to construct a convolution kernel reconstruction module that uses residual structure to improve the encoder decoder, so that the encoder retains more detailed information and the decoder recovers information more effectively, thereby enhancing the expression ability of liver tumor features. Then, to enrich the transmission of global contextual information, a three branch spatial pyramid pooling module is designed to optimize the bottleneck structure of information transmission and break the limitation of a single path. Secondly, a multi-scale feature fusion module is designed to optimize the reuse mechanism of encoder information, enhance the modeling ability of the model for global contextual information, and improve its efficiency in extracting liver tumor features at different scales. Finally, the performance of our method was tested on the LiTS2017 and 3DIRCADb datasets. The experimental results show that our method achieves Dice and IoU values of 97.56% and 95.25% in the liver segmentation task on the LiTS2017 dataset, and 89.71% and 81.58% in the liver tumor segmentation task, respectively. The Dice and IoU values in the liver segmentation task of the 3DIRCADb dataset reached 97.63% and 95.39%, respectively, while the Dice and IoU values in the liver tumor segmentation task reached 89.62% and 81.63%, respectively. This method can effectively alleviate the problem of insufficient expression ability of liver tumor features, and further enhance the model's ability to capture global contextual information.

Prediction of IL-6 Inducing Peptides Based on Graph Neural Networks

CAO Ruifen, HU Weiling, LI Qiangsheng, et al

2025, 53(5):  109-117.  doi:10.12141/j.issn.1000-565X.240242

Asbtract ( 42 )  

Related Articles | Metrics

Interleukin-6 (IL-6) is a highly multifunctional glycoprotein factor that can regulate both innate and adaptive immunity and various aspects of metabolism, including glycolysis, fatty acid oxidation, and oxidative phosphorylation. Many studies have shown that the expression and release of IL-6 in patients infected with viruses significantly increase, and are positively correlated with the severity of the disease. Therefore, identifying IL-6 inducing peptides and exploring their mechanisms of action is very important for developing immune therapies and diagnostic biomarker for the severity of the disease. Currently, the identification of IL-6 inducing peptides methods mostly use traditional machine learning, that require expert knowledge of the field. Therefore, this study proposes a novel IL-6 inducing peptide prediction method (SFGNN-IL6) based on graph neural networks. The predicted structural information of IL-6 inducing peptides are used to construct relation between the nodes of amino acids. The node features are extracted using one-hot encoding, position encoding, and BLOSUM62 encoding, and graph-represented using the encoded features. Then, graph attention mechanism layers and graph convolutional neural network layers are used as dual channels to separately extract features, considering both the update of node weights and the update of node information. Finally, the two types of features are fused for classification of IL-6 inducing peptides. The experimental results validate the effectiveness of the model proposed in this study.

Power & Electrical Engineering

Quantitative Analysis of Interaction in Sub/Super-synchronous Oscillation of DFIG-based Wind Farm Integrated Into Grid Through MMC-HVDC System

#br#

ZHU Lin, ZHAO Xinyue, ZHONG Danting, et al

2025, 53(5):  118-129.  doi:10.12141/j.issn.1000-565X.240078

Asbtract ( 39 )  

Related Articles | Metrics

To quantify the effects of interactions interaction between control loops of grid-connected fan systems, this paper proposes an analytical method that combines the modal analysis method and the relative gain array. Firstly, a small signal model of the DFIG-based wind farms integrated into grid through MMC-HVDC system is established, and the modal analysis method is used to screen out the sub/super-synchronous oscillations which may be affected by the interaction. Secondly, the relative gain array is introduced to confirm the existence of the interaction, quantify and compare the degree of the interaction between the control loops, it shows that the interaction exists between the rotor side converter control loops in DFIG and the constant V/f control loops in MMC-HVDC. Finally, the influence of the electrical distance and controller parameters on the interaction degree is quantitatively evaluated according to the variation of the relative gain array with the influencing factors, and the correctness of the conclusions is verified by the time domain simulation method.

Energy, Power & Electrical Engineering

Study on the Motion Distribution Characteristics of Regular Tetrahedral Particles in Vertical Uppipe

PENG Deqi, ZHOU Jingqiang, FENG Yuan, et al

2025, 53(5):  130-138.  doi:10.12141/j.issn.1000-565X.240222

Asbtract ( 24 )  

Related Articles | Metrics

In order to analyze the motion and distribution law of tetrahedral particle groups in liquid-solid two-phase flow in vertical uppipe, the effects of particle inlet concentration α(1%、2%、3%、4%、5%) and liquid inlet velocity u(1.0、1.2、1.5、1.8、2.0m/s) on the velocity and relative concentration distribution of particle groups in tubes were simulated based on CFD-DEM coupling method, and the accuracy of the numerical simulation is verified by PIV experiments. The results show that the particle group velocity fluctuates along the axial direction and decreases along the radial direction from the center to the wall of the tube within the range of research parameters. Along the radial direction, the relative concentration of particles follows the double peak law. Higher relative concentration of particles in the center of the tube and near the wall, while lower concentration in the transition region. When the particle inlet concentration is 1% and the liquid inlet flow rate is 2.0m/s, the particle concentration near the pipe wall is the highest.

Power & Electrical Engineering

Fusion Transformer based Segmentation Algorithm for Laser Point Cloud of Distribution Lines

DAI Zhou, LIU Yan, MAO Xianying, et al

2025, 53(5):  139-146.  doi:10.12141/j.issn.1000-565X.240542

Asbtract ( 37 )  

Related Articles | Metrics

This paper proposes a laser point cloud segmentation algorithm for power distribution lines based on a fused Transformer, aiming to enhance the precision and efficiency of segmenting and extracting key modules such as power lines, towers, and insulators. A dual-channel parallel architecture feature extraction module is constructed to separately capture high-frequency and low-frequency features, with low-frequency features using average pooling and a fused Transformer extractor, and high-frequency features using max pooling and an MLP module that includes convolutional layers. The feature vectors from both channels are fused to enhance detail extraction. Incorporating an MLP module further refines feature processing for accurate point cloud target segmentation. Extensive experiments validate the algorithm’s accuracy.The algorithm proposed in this paper has the potential advantages of improving accuracy, enhancing automation, increasing robustness, integrating multi-source data, and reducing costs in UAV inspection.