Journal of Northeastern University(Natural Science)

Risk-Oriented Crowd Navigation Strategy Based on Deep Reinforcement Learning

Yang JIANG, Tian-xiang ZHAO, Ruo-huai SUN, Lei WANG

2025, 46 (12): 1-8. DOI: 10.12068/j.issn.1005-3026.2025.12.20240094

Abstract ( 4 )

HTML ( 0)

PDF (2715KB) ( 1 )

To improve robot freezing and suboptimal performance of traditional navigation methods in the presence of dynamic obstacles， a navigation method based on deep reinforcement learning was proposed. The core of this method lies in its risk perception module and path selection module. The risk perception module calculated the collision probability between the robot and nearby dynamic obstacles in real time， allowing the robot to prioritize avoiding more hazardous obstacles. Concurrently， the path selection module evaluated the “passing ability” of the robot in surrounding areas in real time， guiding the robot to choose safer paths. In comparison experiments with a deep reinforcement learning method that lacks these modules， the proposed method achieved the highest navigation success rate in all simulation test environments， with an improvement rate of up to 11%.

Figures and Tables | References | Related Articles | Metrics

Research on Bi-objective Vehicle-Cargo Matching Problem Considering Carbon Emissions

Min HUANG, Ye-xin DU, Hao YU, Xing-wei WANG

2025, 46 (12): 9-18. DOI: 10.12068/j.issn.1005-3026.2025.20240095

Abstract ( 6 )

HTML ( 0)

PDF (2726KB) ( 1 )

To address the issue of insufficient consideration of carbon emissions in vehicle-cargo matching platform decision-making， a bi-objective vehicle-cargo matching model that considers both carbon emissions and platform revenue was proposed. Firstly， an optimization model was constructed with the objectives of minimizing total carbon emissions and maximizing vehicle-cargo matching platform revenue， with load and time constraints. Secondly， to address the model’s multi-objective and non-deterministic polynomial （NP） hard nature， a multi-objective particle swarm optimization （PSO） algorithm was designed， including encoding rules embedded in feasibility analysis， an adaptive elite retention strategy， and a nonlinear decreasing inertia weight. Comparative results on three large-scale examples demonstrate that the proposed algorithm outperforms the NSPSO algorithm， the improved NSGA-II algorithm， and the multi-objective grey wolf algorithm in terms of convergence and uniformity， and it is superior to the latter two algorithms in terms of runtime. Finally， by analyzing the impact of carbon emissions per truck and consignors’ delivery time requirements on carbon emissions， the proposed algorithm provides management insights for platform decision-making.

Figures and Tables | References | Related Articles | Metrics

Prediabetes Detection Method Based on Multi-scale Analysis of HRV

Hong-ru LI, Tong-tong LI, Kang-kang SHI, Ying-hua YANG

2025, 46 (12): 19-28. DOI: 10.12068/j.issn.1005-3026.2025.20240103

Abstract ( 4 )

HTML ( 0)

PDF (4241KB) ( 1 )

Prediabetes is an important stage of abnormal glucose metabolism in the development of diabetes， and its early diagnosis is crucial for global diabetes prevention and control. To explore non-invasive detection methods for prediabetes， heart rate variability （HRV） signals were utilized. By introducing a multi-scale analysis strategy， the global information of the signals was revealed， as well as subtle but important changes at different scales. The CatBoost algorithm was used for classification task. The results show that this method achieves an accuracy of 88.52%， a sensitivity of 83.40%， a specificity of 91.82%， a precision of 86.73%， and an F1-score of 87.40% on the dataset. This study provides a new approach for the diagnosis of prediabetes. The results are especially suitable for wearable devices， offering a potential solution for daily self-health monitoring and disease prevention.

Figures and Tables | References | Related Articles | Metrics

An Improved Small Object Detection Model Based on YOLOv8 for UAV Vision

Ji-hong LIU, Rui-rui SHI

2025, 46 (12): 29-37. DOI: 10.12068/j.issn.1005-3026.2025.20240116

Abstract ( 4 )

HTML ( 0)

PDF (10174KB) ( 0 )

In view of easy false detection and missed detection of small objects in unmanned aerial vehicle （UAV） aerial images， as well as the requirements for real-time performance and lightweight design in UAV detection tasks， an improved lightweight and efficient model based on YOLOv8 was proposed. Firstly， the Neck part of YOLOv8 was simplified into a feature pyramid network， enabling the model to effectively utilize the detailed information extracted by shallow networks. Meanwhile， a feature fusion module was added to provide more favorable features for small object detection to the Head layer. Secondly， an efficient local attention （ELA） mechanism was integrated into the Backbone part to achieve accurate localization of target regions. Experimental results show that compared with YOLOv8s， the parameters and model size of the improved model are reduced by 50%， while the mAP_0.5 and detection speed are improved by 4%. This improved model provides a new idea for the deployment of UAV detection.

Figures and Tables | References | Related Articles | Metrics

Ship Target Detection in Complex Scenarios Based on Improved YOLOv8 Algorithm

Xiao-chen CHE, Shu-hua MA, Ze-xu GUO, Xiao-peng SHA

2025, 46 (12): 38-47. DOI: 10.12068/j.issn.1005-3026.2025.20240118

Abstract ( 4 )

HTML ( 0)

PDF (3025KB) ( 0 )

To improve the accuracy and robustness of ship target detection in complex scenarios， modifications were implemented based on the YOLOv8 algorithm. The CD3 module was introduced in the backbone layer with the parameter-free attention SimAM module embedded. The attention-based scale sequence fusion （ASF） module was incorporated in the neck layer， and an additional detection head was added to the head layer for prediction output. Pruning was adopted to reduce the computations of the model， followed by distillation to further improve model performance. The experiment was conducted on the complex scenario ship detection dataset from Alibaba Tianchi for verification. The results demonstrate that compared with YOLOv8， the improved model achieves increases of 4.7% in AP₅₀ and 2.9% in AP， respectively. Recall and precision are improved by 3.2% and 4.2%， while model parameters and computations are reduced by 56.1% and 30.5%. The optimized model thus improves overall performance while reducing parameters.

Figures and Tables | References | Related Articles | Metrics

Research on Positioning Accuracy Compensation of 6-DoF Robots Based on ISSA-LSSVR

Hua-yu YU, Wen-fu ZHU, Bo XIN, Jun-feng SUN

2025, 46 (12): 48-56. DOI: 10.12068/j.issn.1005-3026.2025.20240136

Abstract ( 4 )

HTML ( 0)

PDF (3321KB) ( 0 )

To enhance the positioning accuracy of six-degree-of-freedom （6-DoF） robots， a method for predicting and compensating for positioning errors of the 6-DoF robot was proposed. A hierarchical line-by-line sampling strategy in the high-frequency workspace of the robot was introduced， and a cumulative measurement error correction formula was established to improve measurement accuracy. Experimental results demonstrate that the robot’s working position significantly affects absolute errors. To address this issue， an error compensation model based on an improved sparrow search algorithm-optimized least squares support vector regression （ISSA-LSSVR） was developed to predict and correct the robot’s inherent positioning errors. The results indicate that， relative to the support vector regression （SVR） algorithm， least squares support vector regression （LSSVR） algorithm， and sparrow search algorithm-optimized LSSVR （SSA-LSSVR）， the ISSA-LSSVR algorithm achieves superior compensation performance. Specifically， the absolute error is reduced by 65.68%， and the maximum error is decreased by 68.95%.

Figures and Tables | References | Related Articles | Metrics

Experimental Study on Hardness of SiC Ceramic Particles-Reinforced High-Entropy Alloy by Laser Cladding

Xue-long WEN, Guang-sheng ZHU, Wen-bo ZHANG, Feng-bing HAN

2025, 46 (12): 57-65. DOI: 10.12068/j.issn.1005-3026.2025.20240137

Abstract ( 5 )

HTML ( 0)

PDF (3217KB) ( 0 )

The ceramic particle-reinforced high-entropy alloy was prepared by laser cladding forming technology. The hardness of the specimens was measured using an electronic Vickers hardness tester， and the effects of laser process parameters， reinforcement phase content of SiC ceramic particles， and Al element content on the hardness of the high-entropy alloy by laser cladding were analyzed. The results indicate that the hardness value of the high-entropy alloy’s cladding layer does not show a significant change with the increase in laser power. With the increase in scanning speed and powder feeding rate， there is a gradual upward trend in the hardness of the cladding layer. With the increase in the reinforcement phase content of SiC ceramic particles， lattice distortion occurs in the specimen； the micro-stress increases， and the hardness of the high-entropy alloy increases obviously. Al element has an obvious regulation effect on the properties of ceramic particle-reinforced high-entropy alloy and can further improve the hardness of the alloy.

Figures and Tables | References | Related Articles | Metrics

Mechanism Analysis and Application of Human-Machine Co-driving Effect Under Severe Driving Conditions

Xian-bin WANG, Wen-long BAO, Lin LI, Ying-zhe ZHANG

2025, 46 (12): 66-77. DOI: 10.12068/j.issn.1005-3026.2025.20249031

Abstract ( 4 )

HTML ( 0)

PDF (3657KB) ( 0 )

The mechanism for the co-driving effect （i.e.， the accuracy of vehicle path tracking and lateral stability） during human-machine co-driving under severe driving conditions is not systematic and insufficient. To address this issue， a human-machine co-driving model was established that incorporated driver state based on the dual-driving dual-control tandem human-machine co-driving framework. Moreover， a method for human-machine co-driving based on model prediction control （MPC） was designed. An in-depth study was conducted on the influence of the driver’s preview time， vehicle speed， road adhesion coefficient， driver state parameters， and controller parameters on the path tracking accuracy and lateral stability of human-machine co-driving vehicles. The results show that the vehicle path tracking accuracy and lateral stability control of human-machine co-driving vehicles under severe working conditions are mutually constrained， and a stronger ability of human-machine co-driving vehicles to resist the perturbation of factors affecting the co-driving effect means a better performance of the designed human-machine co-driving vehicle controller.

Figures and Tables | References | Related Articles | Metrics

Preparation of Polycrystalline Diamond Micro End Mill and Experiment of Its Micro-milling Process for Sapphire

Si-qian GONG, Yao SUN, Si-hui LI

2025, 46 (12): 78-84. DOI: 10.12068/j.issn.1005-3026.2025.20249032

Abstract ( 6 )

HTML ( 0)

PDF (2926KB) ( 1 )

The polycrystalline diamond single-edged helical micro end mill with a diameter of less than 1 mm was fabricated using electrical discharge turning technology featuring non-contact and independence from material strength and hardness limitations. The intrinsic relationship of the tool structure of the polycrystalline diamond spiral micro end mill by electrical discharge turning relative to the feed speed and rotational speed was established， as well as the morphological simulation model of the micro end mill， which successfully solved the difficulties in preparing a polycrystalline diamond helical micro end mill with the diameter of less than 1 mm in the actual machining. The prepared polycrystalline diamond micro end mill was used to carry out experimental research on the micro-milling process of sapphire， and the influence of micro-milling parameters on the three-dimensional surface roughness S_a， S_z， and the triaxial micro-milling force of sapphire was investigated. Moreover， the surface quality， micro-milling force， and tool wear were analyzed to evaluate the micro-milling performance of the polycrystalline diamond micro end mill prepared by electrical discharge turning. The results indicate that the surface roughness S_a of the sapphire processed by the polycrystalline diamond single-edged helical micro end mill can be stably controlled within the range of 0.76~1.00 μm， and the tool wear mode is primarily characterized by damage to the bottom surface of the micro end mill.

Figures and Tables | References | Related Articles | Metrics

Simulation and Experimental Verification of Grinding Forces in Profile Grinding of 18CrNiMo7-6 Steel

Yin-xia ZHANG, Lan-fu LIANG, Zuo-peng SONG, Meng-qi LI

2025, 46 (12): 85-93. DOI: 10.12068/j.issn.1005-3026.2025.20249038

Abstract ( 4 )

HTML ( 0)

PDF (2364KB) ( 0 )

In order to investigate the influence of process parameters on the grinding force in V-notch forming during cylindrical grinding of 18CrNiMo7-6 carburized steel， a cubic boron nitride （CBN） grinding wheel abrasive particle model composed of random polyhedra generated by cutting regular hexahedra with spatial random planes was constructed based on the machining process of fatigue specimens. A three-dimensional simulation model of V-notch forming during cylindrical grinding was established by using ABAQUS finite element simulation software. With the grinding wheel speed n_s， workpiece speed n_w， and radial feed speed v_r as independent variables， single-factor experiments were conducted to study the variation patterns of normal grinding force F_n and tangential grinding force F_t. The validity of the V-notch forming simulation model during plunge grinding was verified through forming grinding force experiments. The results indicate that the normal grinding force is consistently greater than the tangential grinding force， and compared with the workpiece speed and the radial feed speed， the influence of grinding wheel speed on grinding force is more significant. The simulation results agree well with experimental data， where the average errors for normal grinding force predicted by simulations of n_s， n_w， and v_r are 11.06%， 9.21%， and 10.77%， respectively， and those for tangential grinding force are 9.89%， 13.89%， and 15.55%， respectively.

Figures and Tables | References | Related Articles | Metrics

Flow Field Evolution in Spiral Concentrator and Separation Index Prediction of Hematite and Quartz with Different Particle Sizes

Qian WANG, Shu-ling GAO, Xiao-hong ZHOU, Chun-yu LIU

2025, 46 (12): 94-103. DOI: 10.12068/j.issn.1005-3026.2025.20240128

Abstract ( 3 )

HTML ( 0)

PDF (5905KB) ( 0 )

The hydrodynamic characteristics serve as fundamental factors in determining the gravity separation effect. Numerical simulations were used to systematically investigate the evolution of fluid dynamics parameters along the longitudinal travel in a $ϕ$ 400 mm spiral concentrator. For the two feeding systems comprising 90 μm hematite with 38 μm quartz and 59 μm hematite with 38 μm quartz，the variations in particle distribution， migration behavior， and separation efficiency were analyzed. The results indicate that the morphology of the flow film， the velocity distribution， and the intensity of the secondary circulation change significantly within the travel of the first turn， and it takes longitudinal travel of 2~3 turns to stabilize； the stabilization travel is positively correlated with the radial position. Hematite and quartz gradually develop a selective distribution， and the maximum separation efficiency improves overall with the travel. The distribution region of 90 μm hematite is more toward the inner edge， and its migration amount reaches the equilibrium in the 2nd turn； the maximum separation efficiency obtained is 82.16%. The distribution region of 59 μm hematite is more outward， and the migration equilibrium is not reached until the 3rd turn. The maximum separation efficiency obtained in the 2nd and 3rd turns is about 6% lower than that of the 90 μm hematite system. The motion behavior of coarse-grained hematite has an obvious correlation with the evolution of the flow film and velocity distribution. However， the motion of fine-grained hematite exhibits a certain degree of randomness. Consequently， it is necessary to extend the travel or adjust related structural parameters to acquire a satisfactory technical index.

Figures and Tables | References | Related Articles | Metrics

Rheological Properties and Diffusion Mechanism of Cement-Based Slurries with Different Water-Cement Ratios

Lian-chong LI, Yi-teng WANG, Wen-qiang MU, Hong-lei LIU

2025, 46 (12): 104-115. DOI: 10.12068/j.issn.1005-3026.2025.20240114

Abstract ( 4 )

HTML ( 0)

PDF (5849KB) ( 0 )

In order to study the diffusion mechanism of grouting materials in geotechnical and mining engineering， the rheological properties of three types of cement-based slurries under five water-cement ratios were analyzed. Based on the time-dependent viscosity of the slurry and the characteristics of fractured rock masses， a theoretical calculation model for grouting slurry diffusion was established to elucidate the diffusion mechanisms of different cement-based slurries. The results indicate that the slurry of the nano-silica sol composite system has a relatively low density（1.25~1.35 g/cm³）and high fluidity （260~425 mm）. The slag-fly ash and graphene oxide composite systems exhibit low bleeding rates（ $<$ 30%）and good stability. The viscosity of the slurry demonstrates time-dependent characteristics， with a fitting function（ $η (t) = η 0 + k a e k b t$ ）. Time-dependent characteristics weakens as the water-cement ratio increases. The diffusion radii of the three slurries are obtained through computational procedures， revealing that grouting pressure， slurry type， and water-cement ratio collectively influence the diffusion. Under the same diffusion distance， grouting pressure is positively correlated with the time-dependent viscosity characteristics. The graphene oxide composite system requires the least grouting pressure at a given distance， making it easier to diffuse.

Figures and Tables | References | Related Articles | Metrics

Prediction Model of Blasting Fragmentation Based on ACO-BP Neural Network and Its Application

Qing-lei YU, Jia-wei WU, You LI, Jiang-yong PU

2025, 46 (12): 116-123. DOI: 10.12068/j.issn.1005-3026.2025.12.20240122

Abstract ( 4 )

HTML ( 0)

PDF (1478KB) ( 0 )

Blasting fragmentation is governed by the combined interplay of rock properties， blasting parameters， and explosive characteristics. Its accurate prediction is the key to achieving the coordinated optimization of precision blasting in mines and energy consumption during mining and processing. To improve prediction accuracy， a blasting fragmentation prediction model was constructed using ant colony optimization （ACO） and a back propagation （BP） neural network， and the key influencing factors of blasting fragmentation were identified. Based on the Sijiaying open pit mine， a blasting case database was established to train the model and improve prediction accuracy， and the weights of the influencing factors were analyzed. Results show that ACO significantly improves model performance. Among the influencing factors of blasting fragmentation， blasthole spacing has the highest weight， and minimum burden has the lowest weight. Blasting parameters exhibit optimal ranges for blasting fragmentation， but single-parameter adjustments cannot sustainably improve blasting fragmentation. This model provides an effective means and theoretical basis for inverse optimization of blasting design based on blasting effect requirements.

Figures and Tables | References | Related Articles | Metrics

Study on Gas Production Characteristics and Kinetics of Coal and Biomass Co-gasification Under CO₂ Atmosphere

Xi-wen YAO, Hong-bai ZHANG, Kai-li XU, Lei-lei ZHANG

2025, 46 (12): 124-131. DOI: 10.12068/j.issn.1005-3026.2025.20240150

Abstract ( 4 )

HTML ( 0)

PDF (2199KB) ( 0 )

A controlled-atmosphere tube furnace was used to study the gas production and reaction kinetics of coal and biomass during gasification under CO₂ atmosphere. It is found that the pyrolysis of coal and biomass under N₂ atmosphere is divided into three stages： dehydration （30~210 ℃）， rapid pyrolysis （210~400 ℃）， and slow pyrolysis （400~1 000 ℃）. Gasification under CO₂ atmosphere is divided into four stages： dehydration （30~210 ℃）， volatile release （210~400 ℃）， slow weight loss （400~660 ℃）， and Boudouard reaction （660~1 000 ℃）. The gasification conversion rate under CO₂ atmosphere is about 85%， which is higher than that under N₂ atmosphere （53%）. The gasification reaction under CO₂ atmosphere requires higher activation energy than that under N₂ atmosphere， but the reaction rate is faster. The gases produced by the gasification of coal and biomass under CO₂ atmosphere mainly include CO， CO₂， CH₄， and H₂. At 800 ℃， CO gas is regenerated， indicating that CO₂ has a Boudouard reaction with carbon as an oxidant at high temperatures. CO₂ inhibits the formation of CH₄， H₂， and C _m H _n， which contain hydrogen and gases with high calorific values.

Figures and Tables | References | Related Articles | Metrics

Morphology and Elemental Composition of PM_2.5 Single Particle from Road Mobile Sources

Xiu-yan ZHOU, Hui ZHOU, Yu-tao GAO, Wen-hua WANG

2025, 46 (12): 132-140. DOI: 10.12068/j.issn.1005-3026.2025.20240158

Abstract ( 3 )

HTML ( 0)

PDF (2316KB) ( 0 )

Exhaust emissions from motor vehicles and dust stirred up while driving are important components of aerosols， which have a significant impact on the environment and human health. To examine the physicochemical properties of particulate matter associated with motor vehicles， single particle samples of aerosols were collected from a main road with heavy traffic flow （point Ⅰ） and a campus road a certain distance away from the main road （point Ⅱ） in Qinhuangdao City. The morphology and elemental composition of the particles were analyzed by a scanning electron microscopy with energy dispersive X-ray spectroscopy （SEM-EDX）. The results show that the particle number fraction of dust particles at point Ⅰ is 47.1%， much higher than that at point Ⅱ （27.1%）， which mainly comes from motor vehicles’ exhaust. The statistics show that the particle number fraction of spherical fly ash particles in the non-carbonaceous particles at point Ⅰ is 8.0%， which is higher than 2.5% at point Ⅱ， which may come from motor vehicles’ exhaust. Among all the irregular mineral particles， Si-rich particles at both points account for the largest particle number fraction （51.8% at point Ⅰ； 62.1% at point Ⅱ）， possibly indicating a source from surface dust. However， the particle number fraction of Fe-dominant particles at point Ⅰ （13.4%） is significantly higher than that at point Ⅱ （6.3%）. Fe-dominant particles produced by the wear of motor vehicles’ brakes may be an important source of Fe in the atmosphere.

Figures and Tables | References | Related Articles | Metrics

Current Issue