网络首发
Enhancement of N2O adsorption by porous carbon doped with different transition metals
Liu Zheng;Yao Yutong;Zhu Yongzhao;Huang Xiaohui;Wu Xutong;Wu Peilian;Sun Xiang;This study aims to provide an efficient utilization method of waste biomass resources while achieving the adsorption control strategies for greenhouse gases. In this work， the fast-growing eucalyptus bark waste， an abundant forestry byproduct in Guangxi China， was adopted as the carbon precursor. A series of modified porous carbons were fabricated via impregnation with different mass ratios of silver nitrate (AgNO3)， nickel chloride hexahydrate (NiCl2·6H2O)， and ferric chloride hexahydrate (FeCl3·6H2O)， respectively， and followed by high-temperature pyrolysis and activation. The pore structure characteristics and surface functional group of the as-prepared porous carbons were comprehensively characterized through various analytical methods. The effect of pore structure regulation and surface chemistry modulation of different transition-metal doped porous carbon on dynamic N2O adsorption behavior and N2O/CO2 selective adsorption performance were systematically investigated. Combined with kinetic model fitting and apparent sensitivity analysis， the key mechanisms for improving the enhanced N2O adsorption performance were profoundly clarified. The results showed that compared with the undoped porous carbon， all three transition metals doping effectively improved the pore structure and surface chemical properties of the porous carbons， thus facilitating N2O molecules diffusion within interconnected pore channels and promoting interfacial binding affinity between N2O molecules and metal active sites， thereby remarkably enhancing N2O adsorption performance. Moreover， the dynamic adsorption model presents excellent agreement with the experimentally N2O adsorption breakthrough curves， and reliably describes the N₂O kinetic adsorption behavior of the as-prepared porous carbons. Among them， the sample of NPC-Ni-0.5 exhibited the greatest adsorption capacity and the longest breakthrough time among all samples， indicating a pronounced improvement of Ni doping on N2O capture. In contrast， the NPC-Ag-0.5 sample demonstrated the optimal dynamic N2O adsorption selectivity in the N2O/CO2 binary mixed system， implying the distinct advantage of Ag doping for selective separation. Additionally， apparent sensitivity analysis further confirmed the significant contribution of transition-metal associated functional groups to N2O adsorption， and identified the crucial features affecting adsorption performance.
Study on the influence of crystalline state of iPP/PPR melt-blown fibers on adsorption performance of light hydrocarbons
Wang Beifu;Han Leyao;Qi Wenxing;Nie Lihong;This study aims to investigate the effect of crystalline state on the adsorption performance of light hydrocarbons using iPP/PPR melt-blown fibers， and to develop high-efficiency polypropylene-based adsorbents for C₃–C₅ light hydrocarbon volatile organic compounds. A series of composite melt-blown fibers were prepared via melt-blown spinning technology by blending isotactic polypropylene (iPP) with polypropylene random copolymer (PPR) in five mass ratios: 100:0， 60:40， 50:50， 40:60， and 30:70. Differential Scanning Calorimetry (DSC)， X-ray Diffraction (XRD)， and Scanning Electron Microscopy (SEM) were employed to characterize the crystallization behavior， crystal structure， and microscopic morphology of all fiber samples. Dynamic adsorption tests were conducted to obtain breakthrough curves and calculate the saturated adsorption capacity of each sample. Static adsorption experiments at 25 °C were performed across multiple concentration gradients， with the experimental data fitted to both Langmuir and Freundlich adsorption isotherm models. The results indicate that the incorporation of PPR disrupts the regular arrangement of iPP molecular chains and restricts crystal growth. The overall crystallinity of the blended system initially decreases and then increases with rising PPR content. Pure isotactic polypropylene melt-blown fiber exhibits extremely poor adsorption capacity for light hydrocarbons; however， the composite fiber achieves a maximum adsorption capacity of 44.82 mg/g at a PPR mass fraction of 50%. Excessive PPR content leads to flattened and fractured fibers， blocking gas mass transfer channels and reducing adsorption performance. The adsorption process aligns well with the Freundlich model， indicating that the fiber surface presents a heterogeneous adsorption environment. Coupled with analyses of Hansen solubility parameters， adsorption tests under varying partial pressures of light hydrocarbons， and assessments of fiber diameter variations， it is confirmed that the adsorption follows the solution-diffusion mechanism. The amorphous regions of polypropylene predominantly drive the adsorption behavior， while the crystalline regions primarily function to support and maintain the integrity of the fiber network structure. In conclusion， optimizing the two-component ratio can effectively balance the crystallinity and structural integrity of melt-blown fibers. The optimal mass ratio of iPP to PPR is identified as 50/50， significantly enhancing the material's adsorption capacity for light hydrocarbons. This research provides a theoretical framework for the structural design and performance optimization of polypropylene melt-blown fiber adsorbents.
Fire and smoke detection model of wildland-urban interface based on ViT–YOLOv8
Wang Zhe;Tao Mengqi;Zheng Chen;Li Ruiqin;To accurately detect fire and smoke in the complex disturbed environments of the wildland-urban interface， this paper proposes an enhanced detection model based on YOLOv8， named ViT–YOLOv8. This model introduces a Re-Parameterized Vision Transformer (RepViT) module into the backbone network of YOLOv8. Unlike traditional convolutional networks that mix spatial and channel information， this module processes spatial feature extraction and channel feature fusion separately， allowing these two processes to operate independently without interference. This enhances the model’s ability to capture and utilize key features more effectively. Additionally， by leveraging structural re-parameterization， the multi-branch topology of the token mixer is merged into a single depthwise convolution layer， optimizing the allocation of computational resources and improving operational efficiency. Furthermore， a Bottleneck Transformers Network (BoTNet) module integrated with a Ghost Convolutional layer (GhostConv) is incorporated after the spatial pyramid pooling layer in the backbone network. GhostConv increases the number of channels at a low computational cost， enabling the extraction of critical information from the original feature maps. A multi-head self-attention mechanism is introduced into the last three convolutional layers， allowing the model to consistently focus on crucial contextual information within feature maps under complex environmental conditions. This significantly enhances the model’s anti-interference capability while maintaining relatively low computational complexity. In addition to collecting high-quality images of fires in the wildland-urban interface， images of strong light sources that could be mistaken for fires were also gathered as interference samples， simulating the complex lighting conditions typical of such regions and improving the model's robustness. The constructed dataset comprises 4 268 images annotated into three categories: fire， smoke， and other. After training for 650 epochs on this dataset， comparative experiments were conducted with the baseline model YOLOv8n. Experimental results demonstrate that， compared to YOLOv8n， the proposed model achieved improvements in mean average precision (MP)， precision (P)， and recall (R)， increasing from 96.8%， 84.4%， and 97.5% to 97.7%， 84.9%， and 98.0%， respectively. Moreover， ViT–YOLOv8 not only maintained high detection accuracy for fire and smoke targets but also exhibited superior precision in identifying interference items， showcasing excellent anti-interference capability. The model improves both global information capture and local detail recognition， facilitating more accurate detection of fires and their early smoke signs in the complex environment of the wildland-urban interface.
Metro network efficiency analysis considering traction power supply system failures
Li Jie;Zhao Xinyang;Luo Ying;Yin He;Tang Wenpeng;Zhou Suhua;A coupled model of the metro operating network and the traction power supply network is developed to analyze the degradation of metro operating efficiency resulting from power supply failures. The metro network is represented as a network of metro stations and edges， while the power supply network is modeled as a network of power supply nodes and sections. Based on the mapping between power supply sections and metro edges， a cross-network fault propagation mechanism is established. The relationship between the capacity degradation of power supply nodes and the failure of power supply sections is characterized using supply-demand theory. When the available traction power becomes insufficient， train speeds on the affected edges are calculated using a power-speed mapping function， and the corresponding travel times are updated to reflect travel costs under fault conditions. The Stochastic User Equilibrium (SUE) assignment method simulates passenger route choices， taking into account travel time and other related factors. Passenger-flow-weighted network efficiency is employed as the performance indicator for the system; unreachable Origin-Destination (OD) flows are treated as lost demand and excluded from the network efficiency calculation under fault conditions. Three classes of failure scenarios are designed based on the characteristics of traction power supply faults: successive power supply section failures， successive power supply node failures， and independent random failures generated through Monte Carlo simulation. A local metro network in a specific city is selected as a case study， applying the model using smart-card OD data. Experiments under the successive power supply section failure scenario identify critical sections that significantly impact network efficiency， with efficiency losses exceeding 30% during peak hours. Similarly， experiments under the successive power supply node failure scenario identify critical nodes， which induce an average daily efficiency loss of approximately 27%. Both failure scenarios exhibit pronounced temporal characteristics， indicating that the metro network is significantly more sensitive to power supply failures during peak hours than during off-peak hours. Monte Carlo analyses further reveal the system's high vulnerability to spatially clustered failures and independent random failures. Independent random failures can reduce relative metro network efficiency by more than 80% during peak hours， while spatially clustered failures yield a higher mean relative efficiency but involve a substantial risk of abrupt efficiency collapse. The proposed method facilitates the identification and ranking of critical power supply units and quantifies the fluctuations in metro network efficiency under various failure scenarios and demand levels， providing theoretical support for maintenance prioritization and operational planning during peak hours.
Evolution of automobile fires in semi-confined spaces under dynamic ventilation conditions
Liu Changchun;Xu Cheng;Cao Liu;Xi Haoyue;Xin Shijie;Xin Ji;This study investigates the fire characteristics and smoke spread patterns of automobile fires in semi-confined spaces under dynamic ventilation conditions， aiming to explore key parameters for ventilation control. An integrated methodology combining full-scale experiments and numerical simulations was employed. A suffocation and intermittent ventilation test involving two sedans was conducted in a semi-elliptical tunnel (30 m long， 7.6 m wide， and 7.6 m high). Concurrently， a detailed numerical model was developed using the Fire Dynamics Simulator， incorporating the pyrolysis processes of multiple automotive combustible materials (e.g.， foam， fabric， plastics). The model's reliability was validated against experimental temperature data. The effects of ventilation area (simulated by shutter heights of 30 cm， 90 cm， and fully open)， the number of vehicles (one or two)， and vehicle location on fire development were systematically analyzed using the validated model， which comprised approximately 500 000 grids. Experimental results demonstrated that after complete tunnel closure， the fire could be effectively suppressed within approximately 3 minutes due to oxygen deprivation， while smoke rapidly accumulated， filling the entire space within 5 minutes， with a maximum ceiling temperature of 350 ℃. Intermittent ventilation emerged as a critical factor: a small ventilation opening was insufficient to prevent significant fire rekindling， whereas a larger opening introduced sufficient fresh air， leading to vigorous fire rekindling or re-intensification. Numerical simulations further revealed that the maximum heat release rate for two burning cars could reach 14 MW， with the peak consistently occurring within the first 10 minutes after ignition. The maximum ceiling temperatures directly above single and double car fires reached 275 °C and 450 ℃， respectively. Additionally， the closer the vehicles were to the closed end of the tunnel， the more significant the accumulation of hot smoke beneath the ceiling. In conclusion， this study elucidates the coupling mechanism between dynamic ventilation operations and automobile fire development in semi-confined spaces， confirming the existence of a critical ventilation threshold that triggers loss of fire control. The findings underscore the necessity for careful evaluation of ventilation timing and area during emergency responses in similar environments， such as underground parking garages and tunnels， providing quantitative references for fire protection design and firefighting tactics.
Research on carbon sequestration capacity of roadside vegetation based on feature factor response
Zhang Peng;Zhang Yunfei;Wang Ye;Liang Lü;Wang Taoli;Wu Biao;A carbon sequestration calculation model was developed based on roadside vegetation data， using tree height， Diameter at Breast Height (DBH)， and crown spread as primary characteristic factors， while soil organic carbon and root biomass served as secondary factors. Pearson's correlation analysis was conducted to examine the relationships between these characteristic factors and carbon stocks. By combining this analysis with field plot survey data and allometric growth equations， the carbon sequestration capacity of Juniperus chinensis within a 20-meter radius on either side of the road was estimated. The results indicate that the carbon sequestration capacity of vegetation is influenced by multiple factors， including community composition， tree species diversity， tree height， DBH， and crown spread. Specifically， trees with certain dimensions (DBH > 0.18 m， tree height > 6 m， crown spread > 11 m) demonstrated significantly greater carbon sequestration potential. Based on growth stages， the study sample was categorized into three groups: young， middle-aged， and mature stands. During the young stand stage， crown spread exhibited the strongest influence on carbon density; the middle-aged stage proved crucial for increasing carbon density， while the mature stage saw the highest levels of carbon stored per tree and total carbon storage. The optimal planting density was determined to be 0.25–0.33 trees/m2. Over the two-year construction period of the study section， the vegetation sequestered a cumulative total of 50.60 t/km， resulting in carbon neutrality rates of 8% and 5% in the first and second years， respectively. It is estimated that approximately 32 years will be needed to fully offset construction-related carbon emissions. By establishing a tiered planting system based on the varying carbon sequestration capabilities of different tree species and clearly quantifying engineering parameters， it has been demonstrated that prioritizing tree species with high carbon sequestration potential can enhance the overall carbon sequestration efficiency of the community. This approach not only improves the carbon sequestration efficiency of roadside vegetation but also mitigates the delayed carbon sequestration effect observed in young forests.
Aerosol particle size distribution inversion model based on regularization-modified relaxation method
Wang Haibin;Ge Hongjuan;To address the severe ill-posedness of aerosol particle size distribution inversion under low-pressure conditions (10–70 kPa) in aircraft cargo compartments， this study proposes a Tikhonov regularization modified relaxation iterative algorithm. Low-pressure environments enlarge the condition number of the light energy coefficient matrix， reduce measurement signal-to-noise ratio and aggravate the distortion of inversion results; to restrain such oscillation， second-order to fourth-order differential smoothing constraints are embedded into the relaxation iteration framework. The L-curve combined with grid traversal method is adopted to select the optimal regularization parameter λ=10⁻³， balancing data fitting and solution smoothness. Theoretical complexity analysis indicates that the fourth-order smoothing algorithm only brings slight extra single-iteration overhead， while its convergent iteration number drops by 41%， with total runtime below 0.2 s to meet real-time fire detection demands.This paper systematically compares the proposed regularized relaxation methods with classical Chahine algorithm on spherical R-R unimodal/bimodal/trimodal， log-normal (L-N) distributions， as well as non-spherical ellipsoidal and cylindrical particles. Results show that the fourth-order smoothing method outperforms all contrast approaches in all test conditions. For R-R distributions， its inversion accuracy is increased by 9.84%–24.12% compared with Chahine method. Within 0–5.0 μm particle size range of L-N distribution， fourth-order smoothing achieves the highest overall accuracy; only for ultrafine particles of 0–1.0 μm does Chahine perform slightly better than second and third-order smoothing. When 1.0% measurement noise is added， the fourth-order method maintains an inversion accuracy of 97.22% for ellipsoidal and cylindrical particles， showing strong adaptability to irregular particle shapes.A multi-angle light scattering experimental platform with adjustable low-pressure cabin was built for model verification. Under 0.3 atm low-pressure environment， the relative root-mean-square error (RRMSE) of the fourth-order smoothing method is only 3.8%， much lower than the 5.6% error of traditional Chahine algorithm. Despite minor computing cost increment， the fourth-order smoothing regularization relaxation algorithm possesses outstanding anti-noise ability and high reconstruction precision， which can be widely applied to early fire smoke detection in closed low-pressure aircraft cargo compartments. Future work will adopt GPU parallel acceleration to reduce computational overhead and develop hybrid iteration strategies combining Chahine and fourth-order smoothing to further improve inversion efficiency.
Study on the fire suppression efficiency and key parameters of liquid nitrogen in confined spaces for collection storage
Huang Feng;Meng Lingming;Yuan Chengcheng;Wang Xianchun;Yan Weiming;Liu Xiangfei;Archives， libraries， and similar facilities house numerous confined spaces for storing collections， where fires can spread rapidly and pose a high risk of re-ignition. Traditional water-based or chemical extinguishing systems may cause secondary damage， such as water stains or residue on valuable materials. Liquid Nitrogen (LN₂) presents a promising alternative in these scenarios due to its dual-action extinguishing mechanism of cooling and inerting， along with its clean extinguishing properties. To investigate the efficacy of LN₂ fire suppression in such environments， a full-scale experimental platform (7.0 m × 3.0 m × 2.5 m) was constructed. Nitrogen injection tests were performed using a standard wood crib fire source (0.35 m × 0.35 m × 0.08 m， with a total heat release of approximately 21.6 MJ) to simulate burning archival materials. The study systematically analyzed the effects of injection pipe diameters (DN10， DN15， DN20， DN25) and injection flow rates (25%， 50%， 75%， and 100% of the maximum flow rate of 45 L/min) on suppression performance by monitoring temperature variations and oxygen concentrations. Results demonstrate that LN₂ is highly effective for suppressing incipient fires in confined spaces. Under baseline conditions (using DN20 at a flow rate of 45 L/min over a fire area of 21 m2)， complete flame extinguishment was achieved within 57 seconds， during which the ambient temperature dropped from approximately 400 ℃ to around 60 ℃， with a maximum cooling rate of 28.8 ℃/s. Extinguishment time initially decreased but then increased with larger pipe diameters， with DN20 exhibiting optimal performance—improving efficiency by 15% and 21% compared to DN10 and DN25， respectively. The critical oxygen volume fraction at extinguishment was approximately 12%. When maintaining a fixed DN20 pipe diameter， both excessively high and low flow rates reduced suppression efficacy. Optimal performance was observed at 50% of the maximum flow (22.5 L/min)， which reduced extinguishment time by up to 38% compared to other flow rates and achieved a minimum recorded temperature of 0.1 ℃ in the upper space. Variations in flow rate had a more significant impact on cooling than on oxygen inerting. After suppression， temperatures stabilized around 40 ℃， with no signs of re-ignition， confirming LN₂'s deep cooling capability to eliminate smoldering risks in cellulosic materials. These findings provide a scientific basis for implementing LN₂ fire suppression systems in confined spaces used for collection storage.
Physically constrained dynamic selection of kernel network for leakage feature extraction of natural gas valve chamber
Duan Lixiang;Zhu Jiaqi;Zhao Tao;Sun Bingcai;Li Mosong;In natural gas valve chambers， accurately predicting leakage diffusion is often hindered by the aliasing of global diffusion trends with local temperature gradients， particularly in multi-point leakage scenarios. To address the challenge of extracting diffusion boundary information obscured by this feature aliasing， this paper proposes a Physically constrained Dynamic Selection of Kernel Network (P–DSKNet). First， a Dynamic Selective Kernel (DSK) module is designed to achieve adaptive multi-scale feature extraction. By employing parallel branches with varying kernel sizes (e.g.， 3×3 and 5×5)， this module utilizes a sub-network to predict convolutional kernel weights. A two-layer weighting mechanism is then applied to deeply fuse these kernel weights with channel attention vectors， allowing the network to dynamically filter out redundant information and focus on critical temperature gradient features. Second， the architecture integrates the DSK module as a spatial encoder with a Long Short-Term Memory (LSTM) network to rigorously model temporal dependencies. In the decoding phase， residual connections are strategically incorporated to superimpose the input features onto the decoded outputs， effectively mitigating information loss caused by upsampling and enhancing the transfer of deep semantic features. Third， to ensure physical rationality， a hybrid physically constrained loss function is formulated. By integrating spatial gradient continuity and temporal consistency constraints， the training process reinforces physical correlations， ensuring that the predicted temperature fields adhere to fluid dynamics principles. Experimental validation on Computational Fluid Dynamics (CFD) simulated two-point leakage scenarios indicates that P–DSKNet achieves superior prediction performance compared to Physics-Informed Neural Networks (PINN)， LSTM， Convolutional Neural Network–LSTM (CNN–LSTM)， and Transformer models. Quantitatively， compared to CNN–LSTM and PINN， the proposed method increases the coefficient of determination (R2) by 13.3% and 1.1%， reduces the Mean Squared Error (MSE) by 65.3% and 19.0%， and decreases the Mean Absolute Error (MAE) by 35.4% and 14.3%， respectively. These results confirm that P–DSKNet effectively decouples aliased features and substantially improves the accuracy and physical consistency of leakage diffusion predictions.