The Internet of Vehicles (IoV) has become an important direction in the field of intelligent transportation, in which vehicle positioning is a crucial part. SLAM (Simultaneous Localization and Mapping) technology play...The Internet of Vehicles (IoV) has become an important direction in the field of intelligent transportation, in which vehicle positioning is a crucial part. SLAM (Simultaneous Localization and Mapping) technology plays a crucial role in vehicle localization and navigation. Traditional Simultaneous Localization and Mapping (SLAM) systems are designed for use in static environments, and they can result in poor performance in terms of accuracy and robustness when used in dynamic environments where objects are in constant movement. To address this issue, a new real-time visual SLAM system called MG-SLAM has been developed. Based on ORB-SLAM2, MG-SLAM incorporates a dynamic target detection process that enables the detection of both known and unknown moving objects. In this process, a separate semantic segmentation thread is required to segment dynamic target instances, and the Mask R-CNN algorithm is applied on the Graphics Processing Unit (GPU) to accelerate segmentation. To reduce computational cost, only key frames are segmented to identify known dynamic objects. Additionally, a multi-view geometry method is adopted to detect unknown moving objects. The results demonstrate that MG-SLAM achieves higher precision, with an improvement from 0.2730 m to 0.0135 m in precision. Moreover, the processing time required by MG-SLAM is significantly reduced compared to other dynamic scene SLAM algorithms, which illustrates its efficacy in locating objects in dynamic scenes.展开更多
This work demonstrates experimentally the close relation between return currents from relativistic laser-driven target polarization and the quality of the relativistic laser–plasma interaction for laser-driven second...This work demonstrates experimentally the close relation between return currents from relativistic laser-driven target polarization and the quality of the relativistic laser–plasma interaction for laser-driven secondary sources,taking as an example ion acceleration by target normal sheath acceleration.The Pearson linear correlation of maximum return current amplitude and proton spectrum cutoff energy is found to be in the range from~0.70 to 0.94.kA-scale return currents rise in all interaction schemes where targets of any kind are charged by escaping laser-accelerated relativistic electrons.Their precise measurement is demonstrated using an inductive scheme that allows operation at high repetition rates.Thus,return currents can be used as a metrological online tool for the optimization of many laser-driven secondary sources and for diagnosing their stability.In particular,in two parametric studies of laser-driven ion acceleration,we carry out a noninvasive online measurement of return currents in a tape target system irradiated by the 1 PW VEGA-3 laser at Centro de Láseres Pulsados:first the size of the irradiated area is varied at best compression of the laser pulse;second,the pulse duration is varied by means of induced group delay dispersion at best focus.This work paves the way to the development of feedback systems that operate at the high repetition rates of PW-class lasers.展开更多
The level of ground shaking,as determined by the peak ground acceleration(PGA),can be used to analyze seismic hazard at a certain location and is crucial for constructing earthquake-resistant structures.Predicting the...The level of ground shaking,as determined by the peak ground acceleration(PGA),can be used to analyze seismic hazard at a certain location and is crucial for constructing earthquake-resistant structures.Predicting the PGA immediately after an earthquake occurs allows for the issuing of a warning by an earthquake early warning system.In this study,we propose a deep learning model,ConvMixer,to predict the PGA recorded by weak-motion velocity seismometers in Japan.We use 5-s threecomponent seismograms,from 2 s before until 3 s after the P-wave arrival time of the earthquake.Our dataset comprised more than 50,000 single-station waveforms recorded by 10 seismic stations in the K-NET,Kiki-NET,and Hi-Net networks between 2004 and 2023.The proposed ConvMixer is a patch-based model that extracts global features from input seismic data and predicts the PGA of an earthquake by combining depth and pointwise convolutions.The proposed ConvMixer network had a mean absolute error of 2.143 when applied to the test set and outperformed benchmark deep learning models.In addition,the proposed ConvMixer demonstrated the ability to predict the PGA at the corresponding station site based on 1-second waveforms obtained immediately after the arrival time of the P-wave.展开更多
We put forward a new design of a compact beam transport system for intense laser-driven proton therapy,where instead of using conventional pulsed solenoids,our design relies on a helical coil irradiated by a nanosecon...We put forward a new design of a compact beam transport system for intense laser-driven proton therapy,where instead of using conventional pulsed solenoids,our design relies on a helical coil irradiated by a nanosecond laser pulse to generate strong magnetic fields for focusing protons.A pair of dipole magnets and apertures are employed to further filter protons with large divergences and low energies.Our numerical studies combine particle-in-cell simulations for laser-plasma interaction to generate high-energy monoenergetic proton beams,finite element analysis for evaluating the magnetic field distribution inside the coil,and MonteCarlo simulations for beam transport and energy deposition.Our results show that with this design,a spread-out Bragg peak in a range of several centimeters to a deep-seated tumor with a dose of approximately 16.5 cGy and fluctuation around 2% can be achieved.The instantaneous dose rate reaches up to 10^(9)Gy/s,holding the potential for future FLASH radiotherapy research.展开更多
Background Higher accelerometer-assessed volume and intensity of physical activity(PA)have been associated with a longer life expectancy but can be difficult to translate into recommended doses of PA.We aimed to:(a)im...Background Higher accelerometer-assessed volume and intensity of physical activity(PA)have been associated with a longer life expectancy but can be difficult to translate into recommended doses of PA.We aimed to:(a)improve interpretability by producing UK Biobank age-referenced centiles for PA volume and intensity;(b)inform public-health messaging by examining how adding recommended quantities of moderate and vigorous PA affect PA volume and intensity.Methods 92,480 UK Biobank participants aged 43-80 years with wrist-worn accelerometer data were included.Average acceleration and intensity gradient were derived as proxies for PA volume and intensity.We generated sex-specific centile curves using Generalized Additive Models for Location Scale and Shape(GAMLSS)and modeled the effect of adding moderate(walking)or vigorous(running)activity on the combined change in the volume and intensity centiles(change in PA profile).Results In men,volume was lower as age increased while intensity was lower after age 55;in women,both volume and intensity were lower as age increased.Adding 150 min of moderate PA weekly(5×30 min walking)increased the PA profile by 4 percentage points.Defining moderate PA as brisk walking approximately doubled the increase(9 percentage points)while 75 min of vigorous PA weekly(5×15 min running)trebled the increase(13 percentage points).Conclusion These UK Biobank reference centiles provide a benchmark for interpretation of accelerometer data.Application of our translational methods demonstrate that meeting PA guidelines through shorter duration vigorous activity is more beneficial to the PA profile(volume and intensity)than longer duration moderate activity.展开更多
This study focused on the hydraulic conductivity of sand in centrifuge modeling.A self-designed temperature-controlled falling-head permeameter apparatus was used,and a series of falling-head seepage tests were perfor...This study focused on the hydraulic conductivity of sand in centrifuge modeling.A self-designed temperature-controlled falling-head permeameter apparatus was used,and a series of falling-head seepage tests were performed on sand samples with various porosities at different temperatures and centrifugal accelerations.The objectives were to qualitatively and quantitatively investigate the effects of temperature,porosity,and centrifugal acceleration on the hydraulic conductivity of sand and to study the applicability of the Kozeny-Carman equation for the centrifugal environment.Test results showed that in a similar temperature range and under the same porosity,the hydraulic conductivity of the sand is linearly correlated with centrifugal acceleration.When subjected to the same centrifugal acceleration and in a similar temperature range,the hydraulic conductivity of the sand exhibits an almost linear increase in relation to its porosity function(s^(3)/(1−s)^(2));the functional relationships between the hydraulic conductivity of the sand and temperature,centrifugal acceleration level,and porosity were established using two pathways.When the centrifugal acceleration is less than 50g,the Kozeny-Carman equation is effectively accurate in predicting the hydraulic conductivity of sand;however,when the centrifugal acceleration exceeds 50g,it is important to consider a significant error.展开更多
Hydrodynamic cavitation,as an efficient technique applied in many physical and chemical treatment methods,has been widely used by various industries and in several technological fields.Relevant generators,designed wit...Hydrodynamic cavitation,as an efficient technique applied in many physical and chemical treatment methods,has been widely used by various industries and in several technological fields.Relevant generators,designed with specific structures and parameters,can produce cavitation effects,thereby enabling effective treatment and reasonable transformation of substances.This paper reviews the design principles,performance,and practical applications associated with different types of cavitation generators,aiming to provide theoretical support for the optimization of these systems.It systematically analyzes the underpinning mechanisms and the various factors influencing the cavitation phenomena,also conducting a comparative analysis of the performance of different types of generators.Specific applications dealing with wastewater treatment,chemical reaction acceleration,and other fields are discussed together with the advantages,disadvantages,and applicability of each type of cavitation generator.We also explore research progress in areas such as cavitation stability,energy efficiency,and equipment design upgrades.The study concludes by forecasting the application prospects of intelligent design and computational fluid dynamics(CFD)in optimizing and advancing cavitation generators.It proposes new ideas for the further development of cavitation technology and highlights directions for its widespread future application.展开更多
Alcohol oxidation is a widely used green chemical reaction.The reaction process produces flammable and explosive hydrogen,so the design of the reactor must meet stringent safety requirements.Based on the limited exper...Alcohol oxidation is a widely used green chemical reaction.The reaction process produces flammable and explosive hydrogen,so the design of the reactor must meet stringent safety requirements.Based on the limited experimental data,utilizing the traditional numerical method of computational fluid dynamics(CFD)to simulate the gas-liquid two-phase flow reactor can mitigate the risk of danger under varying working conditions.However,the calculation process is highly time-consuming.Therefore,by integrating process simulation,computational fluid dynamics,and deep learning technologies,an intelligent hybrid chemical model based on machine learning was proposed to expedite CFD calculations,enhance the prediction of flow fields,conversion rates,and concentrations inside the reactor,and offer insights for designing and optimizing the reactor for the alcohol oxidation system.The results show that the hybrid model based on the long and short-term memory neural network achieves 99.8%accuracy in conversion rate prediction and 99.9%accuracy in product concentration prediction.Through validation,the hybrid model is accelerated by about 360 times compared with instrumental analysis in conversion rate prediction and about 45 times compared with CFD calculation in concentration distribution prediction.This hybrid model can quickly predict the conversion rate and product concentration distribution in the gas-liquid two-phase flow reactor and provide a model reference for fast prediction and accurate control in the actual chemical production process.展开更多
This study investigates the paradoxical detonation behavior of TKX-50,a nitrogen-rich energetic material,exhibiting higher detonation velocities but lower metal acceleration ability compared to HMX.Through experimenta...This study investigates the paradoxical detonation behavior of TKX-50,a nitrogen-rich energetic material,exhibiting higher detonation velocities but lower metal acceleration ability compared to HMX.Through experimental measurements and theoretical calculations,we propose a novel three-factor competition mechanism to explain this phenomenon.TKX-50-based PBX formulations achieved detonation velocities up to 9100 m/s,surpassing HMX-based counterparts.However,cylinder expansion tests revealed a 15%reduction in metal acceleration ability.Thermochemical measurements showed lower detonation heat for TKX-50(4900 J/g)versus HMX(5645 J/g).Our mechanism involves:(1)compositional effects prevailing at high pressures;(2)Energy release becoming essential as pressure drops;(3)Pressure-dependent product composition evolution functioning at low pressure.VLW code calculations unveiled a"crossover"in Hugoniot curves,lending support to this mechanism.This study furnishes a new framework for comprehending the performance of nitrogen-rich energetic materials,with significant implications for the design and optimization of future high-energy density materials.展开更多
Both acceleration and pseudo-acceleration response spectra play important roles in structural seismic design.However,only one of them is generally provided in most seismic codes.Therefore,many studies have attempted t...Both acceleration and pseudo-acceleration response spectra play important roles in structural seismic design.However,only one of them is generally provided in most seismic codes.Therefore,many studies have attempted to develop conversion models between the acceleration response spectrum(SA)and the pseudo-acceleration response spectrum(PSA).Our previous studies found that the relationship between SA and PSA is affected by magnitude,distance,and site class.Subsequently,we developed an SA/PSA model incorporating these effects.However,this model is suitable for cases with small and moderate magnitudes and its accuracy is not good enough for cases with large magnitudes.This paper aims to develop an efficient SA/PSA model by considering influences of magnitude,distance,and site class,which can be applied to cases not only with small or moderate magnitudes but also with large ones.For this purpose,regression analyses were conducted using 16,660 horizontal seismic records with a wider range of magnitude.The magnitude of these seismic records varies from 4 to 9 and the distances vary from 10 to 200 km.These ground motions were recorded at 338 stations covering four site classes.By comparing them with existing models,it was found that the proposed model shows better accuracy for cases with any magnitudes,distances,and site classes considered in this study.展开更多
In order to solve the problems of limited performance of single and homogeneous sensors and high false alarm rate caused by environmental noise,heterogeneous sensor data fusion(HSDF),a fall detection framework based o...In order to solve the problems of limited performance of single and homogeneous sensors and high false alarm rate caused by environmental noise,heterogeneous sensor data fusion(HSDF),a fall detection framework based on heterogeneous sensor fusion of acceleration and audio signals,is proposed in this paper.By analyzing the heterogeneity of acceleration data and audio data,the framework uses Dempstershafer Theory(D-S Theory)to integrate the output of acceleration and audio data at decision level.Firstly,a normalized window interception algorithm——anomaly location window algorithm(ALW)is proposed by analyzing the fall process and the characteristics of acceleration changes based on acceleration data.Secondly,the one-dimensional residual convolutional network(1D-ReCNN)is designed for fall detection based on the audio data.Finally,it is verified that the HSDF framework has good advantages in terms of sensitivity and false alarm rate by the collection of volunteers’simulated fall data and free living data in real environment.展开更多
Early exiting has shown significant potential in accelerating the inference of pre-trained language models(PLMs)by allowing easy samples to exit from shallow layers.However,existing early exiting methods primarily rel...Early exiting has shown significant potential in accelerating the inference of pre-trained language models(PLMs)by allowing easy samples to exit from shallow layers.However,existing early exiting methods primarily rely on local information from individual samples to estimate prediction uncertainty for making exiting decisions,overlooking the global information provided by the sample population.This impacts the estimation of prediction uncertainty,compromising the reliability of exiting de-cisions.To remedy this,inspired by principal component analysis(PCA),the authors define a residual score to capture the deviation of features from the principal space of the sample population,providing a global perspective for estimating prediction uncertainty.Building on this,a two-stage exiting strategy is proposed that integrates global information from residual scores with local information from energy scores at both the decision and feature levels.This strategy incorporates three-way decisions to enable more reliable exiting decisions for boundary region samples by delaying judgement.Extensive experiments on the GLUE benchmark validate that the method achieves an average speed-up ratio of 2.17×across all tasks with minimal per-formance degradation.Additionally,it surpasses the state-of-the-art E-LANG by 11%in model acceleration,along with a performance improvement of 0.6 points,demonstrating a better performance-efficiency trade-off.展开更多
Bedding parallel stepped rock slopes exist widely in nature and are used in slope engineering.They are characterized by complex topography and geological structure and are vulnerable to shattering under strong earthqu...Bedding parallel stepped rock slopes exist widely in nature and are used in slope engineering.They are characterized by complex topography and geological structure and are vulnerable to shattering under strong earthquakes.However,no previous studies have assessed the mechanisms underlying seismic failure in rock slopes.In this study,large-scale shaking table tests and numerical simulations were conducted to delineate the seismic failure mechanism in terms of acceleration,displacement,and earth pressure responses combined with shattering failure phenomena.The results reveal that acceleration response mutations usually occur within weak interlayers owing to their inferior performance,and these mutations may transform into potential sliding surfaces,thereby intensifying the nonlinear seismic response characteristics.Cumulative permanent displacements at the internal corners of the berms can induce quasi-rigid displacements at the external corners,leading to greater permanent displacements at the internal corners.Therefore,the internal corners are identified as the most susceptible parts of the slope.In addition,the concept of baseline offset was utilized to explain the mechanism of earth pressure responses,and the result indicates that residual earth pressures at the internal corners play a dominant role in causing deformation or shattering damage.Four evolutionary deformation phases characterize the processes of seismic responses and shattering failure of the bedding parallel stepped rock slope,i.e.the formation of tensile cracks at the internal corners of the berm,expansion of tensile cracks and bedding surface dislocation,development of vertical tensile cracks at the rear edge,and rock mass slipping leading to slope instability.Overall,this study provides a scientific basis for the seismic design of engineering slopes and offers valuable insights for further studies on preventing seismic disasters in bedding parallel stepped rock slopes.展开更多
Purpose–This paper aims to analyze the transverse vibration characteristics of the high speed train window glass when passing through tunnel.Design/methodology/approach–The lateral vibration acceleration response of...Purpose–This paper aims to analyze the transverse vibration characteristics of the high speed train window glass when passing through tunnel.Design/methodology/approach–The lateral vibration acceleration response of glass chamber of high-speed train CR400BF-A on Beijing-Chengdu high-speed railway was tested at different speeds through the tunnel entrance,exit,tunnel interior,Tunnel Group and rendezvous time in the tunnel,the lateral distribution characteristics of vibration frequency and vibration power amplification coefficient of glass of high-speed train were analyzed.Findings–The results show that:The vibration of the high-speed train glass increases significantly during the tunnel,and the amplitude of vibration acceleration in the tunnel is significantly higher than outside the tunnel as the travel speed increases;the amplitude of lateral vibration acceleration of the glass of a high-speed train does not vary with changes in tunnel length and is not affected by the aerodynamic effects of the tunnel when traveling inside the tunnel,but its vibrations create noticeable fluctuations during variations when encountering oncoming traffic;The vibration characteristics of the high-speed train glass are forced harmonic vibrations,the excitation frequency does not vary with travel speed and travel position changes inside and outside the tunnel.The lateral vibration acceleration of the glass of a high-speed train is applied vertically and uniformly to the glass surface as an“inertial force”and creates a cyclic bending vibration stress that can easily lead to fatigue damage.Originality/value–The research results provide guidance for the prevention of glass failure in high-speed trains.展开更多
Computing-in-memory(CIM)has been a promising candidate for artificial-intelligent applications thanks to the absence of data transfer between computation and storage blocks.Resistive random access memory(RRAM)based CI...Computing-in-memory(CIM)has been a promising candidate for artificial-intelligent applications thanks to the absence of data transfer between computation and storage blocks.Resistive random access memory(RRAM)based CIM has the advantage of high computing density,non-volatility as well as high energy efficiency.However,previous CIM research has predominantly focused on realizing high energy efficiency and high area efficiency for inference,while little attention has been devoted to addressing the challenges of on-chip programming speed,power consumption,and accuracy.In this paper,a fabri-cated 28 nm 576K RRAM-based CIM macro featuring optimized on-chip programming schemes is proposed to address the issues mentioned above.Different strategies of mapping weights to RRAM arrays are compared,and a novel direct-current ADC design is designed for both programming and inference stages.Utilizing the optimized hybrid programming scheme,4.67×programming speed,0.15×power saving and 4.31×compact weight distribution are realized.Besides,this macro achieves a normalized area efficiency of 2.82 TOPS/mm2 and a normalized energy efficiency of 35.6 TOPS/W.展开更多
A novel porous shock absorption layer is put forward in this study, and the shock absorption performance of the porous shock absorption layer is evaluated based on three-dimensional pseudo-static analysis. The modifie...A novel porous shock absorption layer is put forward in this study, and the shock absorption performance of the porous shock absorption layer is evaluated based on three-dimensional pseudo-static analysis. The modified reaction acceleration method is adopted and validated in the three-dimensional model. Seven ground motions are selected and the peak ground acceleration is adjusted to 0.2 g, 0.4 g and 0.6 g. The impact of the void ratio and thickness of the porous shock absorption layer is studied, while the surrounding rock grade and tunnel depth are also investigated. The numerical results show that the porous shock absorption layer has good shock absorption performance and can effectively reduce the maximum internal force of the secondary lining, but it cannot reduce the maximum horizontal relative displacement of the secondary lining. The circumferential rubber strip in the porous shock absorption layer will reduce shock absorption performance. The results of parameter analysis indicate that the shock absorption performance of the porous shock absorption layer increases with the increase of the void ratio and thickness, and it has good shock absorption performance under different surrounding rock grades and tunnel depths.展开更多
Variable Cycle Engine(VCE)serves as the core system in achieving future advanced fighters with cross-generational performance and mission versatility.However,the resultant complex configuration and strong coupling of ...Variable Cycle Engine(VCE)serves as the core system in achieving future advanced fighters with cross-generational performance and mission versatility.However,the resultant complex configuration and strong coupling of control parameters present significant challenges in designing acceleration and deceleration control schedules.To thoroughly explore the performance potential of engine,a global integration design method for acceleration and deceleration control schedule based on inner and outer loop optimization is proposed.The outer loop optimization module employs Integrated Surrogate-Assisted Co-Differential Evolutionary(ISACDE)algorithm to optimize the variable geometry adjustment laws based on B-spline curve,and the inner loop optimization module adopts the fixed-state method to design the open-loop fuel–air ratio control schedules,which are aimed at minimizing the acceleration and deceleration time under multiple constraints.Simulation results demonstrate that the proposed global integration design method not only furthest shortens the acceleration and deceleration time,but also effectively safeguards the engine from overlimit.展开更多
Coralline soils,specialized materials found extensively in the South China Sea,are playing an increasingly vital role in engineering projects.However,like most terrigenous soils,fine-grained coral soil is prone to shr...Coralline soils,specialized materials found extensively in the South China Sea,are playing an increasingly vital role in engineering projects.However,like most terrigenous soils,fine-grained coral soil is prone to shrinkage and cracking,which can significantly affect its engineering properties and ultimately jeopardize engineering safety.This paper presents a desiccation cracking test of fine-grained coral soil,with a particular focus on the thickness effect.The study involved measuring the water content and recording the evolution of desiccation cracking.Advanced image processing technology is employed to analyze the variations in crack parameters,clod parameters,fractal dimensions,frequency distributions,and desiccation cracking propagation velocities of fine-grained coral soil.Furthermore,the dynamic evolution of desiccation cracking under the influence of layer thickness is analyzed.A comprehensive crack evolution model is proposed,encompassing both top-down and bottom-up crack propagation,as well as internal tensile cracking.This work introduces novel metrics for the propagation velocity of the total crack area,the characteristic propagation velocities of desiccation cracks,and the acceleration of crack propagation.Through data fitting,theoretical formulas for soil water evaporation,propagation velocities of desiccation cracks,and crack propagation acceleration are derived,laying a foundation for future soil cracking theories.展开更多
This paper introduced a compact high flux polarized neutron beam generator scheme,which used air as the working medium and had low energy consumption.The neutron beam generator adopted a linear three compartment confi...This paper introduced a compact high flux polarized neutron beam generator scheme,which used air as the working medium and had low energy consumption.The neutron beam generator adopted a linear three compartment configuration,sequentially using nitrogen nucleus tandem near range accelerated polarization target spallation nuclear reaction technology,neutron multiplication technology,neutron beam polarization and near range acceleration technology,neutron focusing and shooting control technology.Through design and equivalent verification,it has been proven that the total length of the device does not exceed 5 m,the effective range can reach several hundred kilometers,the neutron flux at the muzzle is not less than 10^(25) n·cm^(-2)·s^(-1),which attenuates to 10^(10) n·cm^(-2)·s^(-1) at a distance of several 100 km,and this flux can effectively strike the target.It can be used as a defensive directed energy weapon with high energy density and has broad application prospects.展开更多
Lost acceleration response reconstruction is crucial for assessing structural conditions in structural health monitoring(SHM).However,traditional methods struggle to address the reconstruction of acceleration response...Lost acceleration response reconstruction is crucial for assessing structural conditions in structural health monitoring(SHM).However,traditional methods struggle to address the reconstruction of acceleration responses with complex features,resulting in a lower reconstruction accuracy.This paper addresses this challenge by leveraging the advanced feature extraction and learning capabilities of fully convolutional networks(FCN)to achieve precise reconstruction of acceleration responses.In the designed network architecture,the incorporation of skip connections preserves low-level details of the network,greatly facilitating the flow of information and improving training efficiency and accuracy.Dropout techniques are employed to reduce computational load and enhance feature extraction.The proposed FCN model automatically extracts high-level features from the input data and establishes a nonlinearmapping relationship between the input and output responses.Finally,the accuracy of the FCN for structural response reconstructionwas evaluated using acceleration data from an experimental arch rib and comparedwith several traditional methods.Additionally,this approach was applied to reconstruct actual acceleration responses measured by an SHM system on a long-span bridge.Through parameter analysis,the feasibility and accuracy of aspects such as available response positions,the number of available channels,and multi-channel response reconstruction were explored.The results indicate that this method exhibits high-precision response reconstruction capability in both time and frequency domains.,with performance surpassing that of other networks,confirming its effectiveness in reconstructing responses under various sensor data loss scenarios.展开更多
基金funded by the Natural Science Foundation of the Jiangsu Higher Education Institutions of China(grant number 22KJD440001)Changzhou Science&Technology Program(grant number CJ20220232).
文摘The Internet of Vehicles (IoV) has become an important direction in the field of intelligent transportation, in which vehicle positioning is a crucial part. SLAM (Simultaneous Localization and Mapping) technology plays a crucial role in vehicle localization and navigation. Traditional Simultaneous Localization and Mapping (SLAM) systems are designed for use in static environments, and they can result in poor performance in terms of accuracy and robustness when used in dynamic environments where objects are in constant movement. To address this issue, a new real-time visual SLAM system called MG-SLAM has been developed. Based on ORB-SLAM2, MG-SLAM incorporates a dynamic target detection process that enables the detection of both known and unknown moving objects. In this process, a separate semantic segmentation thread is required to segment dynamic target instances, and the Mask R-CNN algorithm is applied on the Graphics Processing Unit (GPU) to accelerate segmentation. To reduce computational cost, only key frames are segmented to identify known dynamic objects. Additionally, a multi-view geometry method is adopted to detect unknown moving objects. The results demonstrate that MG-SLAM achieves higher precision, with an improvement from 0.2730 m to 0.0135 m in precision. Moreover, the processing time required by MG-SLAM is significantly reduced compared to other dynamic scene SLAM algorithms, which illustrates its efficacy in locating objects in dynamic scenes.
基金funding from the European Union’s Horizon 2020 research and innovation program through the European IMPULSE project under Grant Agreement No.871161from LASERLAB-EUROPE V under Grant Agreement No.871124+6 种基金from the Grant Agency of the Czech Republic(Grant No.GM23-05027M)Grant No.PDC2021120933-I00 funded by MCIN/AEI/10.13039/501100011033by the European Union Next Generation EU/PRTRsupported by funding from the Ministerio de Ciencia,Innovación y Universidades in Spain through ICTS Equipment Grant No.EQC2018-005230-Pfrom Grant No.PID2021-125389O A-I00 funded by MCIN/AEI/10.13039/501100011033/FEDER,UEby“ERDF A Way of Making Europe”by the European Unionfrom grants of the Junta de Castilla y León with Grant Nos.CLP263P20 and CLP087U16。
文摘This work demonstrates experimentally the close relation between return currents from relativistic laser-driven target polarization and the quality of the relativistic laser–plasma interaction for laser-driven secondary sources,taking as an example ion acceleration by target normal sheath acceleration.The Pearson linear correlation of maximum return current amplitude and proton spectrum cutoff energy is found to be in the range from~0.70 to 0.94.kA-scale return currents rise in all interaction schemes where targets of any kind are charged by escaping laser-accelerated relativistic electrons.Their precise measurement is demonstrated using an inductive scheme that allows operation at high repetition rates.Thus,return currents can be used as a metrological online tool for the optimization of many laser-driven secondary sources and for diagnosing their stability.In particular,in two parametric studies of laser-driven ion acceleration,we carry out a noninvasive online measurement of return currents in a tape target system irradiated by the 1 PW VEGA-3 laser at Centro de Láseres Pulsados:first the size of the irradiated area is varied at best compression of the laser pulse;second,the pulse duration is varied by means of induced group delay dispersion at best focus.This work paves the way to the development of feedback systems that operate at the high repetition rates of PW-class lasers.
基金the National Research Institute of Astronomy and Geophysics (NRIAG) for supporting this work
文摘The level of ground shaking,as determined by the peak ground acceleration(PGA),can be used to analyze seismic hazard at a certain location and is crucial for constructing earthquake-resistant structures.Predicting the PGA immediately after an earthquake occurs allows for the issuing of a warning by an earthquake early warning system.In this study,we propose a deep learning model,ConvMixer,to predict the PGA recorded by weak-motion velocity seismometers in Japan.We use 5-s threecomponent seismograms,from 2 s before until 3 s after the P-wave arrival time of the earthquake.Our dataset comprised more than 50,000 single-station waveforms recorded by 10 seismic stations in the K-NET,Kiki-NET,and Hi-Net networks between 2004 and 2023.The proposed ConvMixer is a patch-based model that extracts global features from input seismic data and predicts the PGA of an earthquake by combining depth and pointwise convolutions.The proposed ConvMixer network had a mean absolute error of 2.143 when applied to the test set and outperformed benchmark deep learning models.In addition,the proposed ConvMixer demonstrated the ability to predict the PGA at the corresponding station site based on 1-second waveforms obtained immediately after the arrival time of the P-wave.
基金supported by the National Key R&D Program of China(Nos.2022YFA1603200 and 2022YFA1603201)National Natural Science Foundation of China(Nos.12135001,11921006,12475243 and 11825502)+1 种基金Strategic Priority Research Program of CAS(No.XDA25050900)support from the National Natural Science Funds for Distinguished Young Scholar(No.11825502)。
文摘We put forward a new design of a compact beam transport system for intense laser-driven proton therapy,where instead of using conventional pulsed solenoids,our design relies on a helical coil irradiated by a nanosecond laser pulse to generate strong magnetic fields for focusing protons.A pair of dipole magnets and apertures are employed to further filter protons with large divergences and low energies.Our numerical studies combine particle-in-cell simulations for laser-plasma interaction to generate high-energy monoenergetic proton beams,finite element analysis for evaluating the magnetic field distribution inside the coil,and MonteCarlo simulations for beam transport and energy deposition.Our results show that with this design,a spread-out Bragg peak in a range of several centimeters to a deep-seated tumor with a dose of approximately 16.5 cGy and fluctuation around 2% can be achieved.The instantaneous dose rate reaches up to 10^(9)Gy/s,holding the potential for future FLASH radiotherapy research.
基金supported by the National Institute for Health Research (NIHR) Leicester Biomedical Research Centre and NIHR Applied Research Collaboration East Midlands (ARC EM, (IS-BRC-1215-20010))funded by the National Institute for Health and Care Research Bristol Biomedical Research Centre (IS-BRC-1215-20011)supported by a UKRI research grant (EP/X042464/1)。
文摘Background Higher accelerometer-assessed volume and intensity of physical activity(PA)have been associated with a longer life expectancy but can be difficult to translate into recommended doses of PA.We aimed to:(a)improve interpretability by producing UK Biobank age-referenced centiles for PA volume and intensity;(b)inform public-health messaging by examining how adding recommended quantities of moderate and vigorous PA affect PA volume and intensity.Methods 92,480 UK Biobank participants aged 43-80 years with wrist-worn accelerometer data were included.Average acceleration and intensity gradient were derived as proxies for PA volume and intensity.We generated sex-specific centile curves using Generalized Additive Models for Location Scale and Shape(GAMLSS)and modeled the effect of adding moderate(walking)or vigorous(running)activity on the combined change in the volume and intensity centiles(change in PA profile).Results In men,volume was lower as age increased while intensity was lower after age 55;in women,both volume and intensity were lower as age increased.Adding 150 min of moderate PA weekly(5×30 min walking)increased the PA profile by 4 percentage points.Defining moderate PA as brisk walking approximately doubled the increase(9 percentage points)while 75 min of vigorous PA weekly(5×15 min running)trebled the increase(13 percentage points).Conclusion These UK Biobank reference centiles provide a benchmark for interpretation of accelerometer data.Application of our translational methods demonstrate that meeting PA guidelines through shorter duration vigorous activity is more beneficial to the PA profile(volume and intensity)than longer duration moderate activity.
基金supported by the Basic Science Center Program for Multiphase Media Evolution in Hypergravity of the National Natural Science Foundation of China(No.51988101)。
文摘This study focused on the hydraulic conductivity of sand in centrifuge modeling.A self-designed temperature-controlled falling-head permeameter apparatus was used,and a series of falling-head seepage tests were performed on sand samples with various porosities at different temperatures and centrifugal accelerations.The objectives were to qualitatively and quantitatively investigate the effects of temperature,porosity,and centrifugal acceleration on the hydraulic conductivity of sand and to study the applicability of the Kozeny-Carman equation for the centrifugal environment.Test results showed that in a similar temperature range and under the same porosity,the hydraulic conductivity of the sand is linearly correlated with centrifugal acceleration.When subjected to the same centrifugal acceleration and in a similar temperature range,the hydraulic conductivity of the sand exhibits an almost linear increase in relation to its porosity function(s^(3)/(1−s)^(2));the functional relationships between the hydraulic conductivity of the sand and temperature,centrifugal acceleration level,and porosity were established using two pathways.When the centrifugal acceleration is less than 50g,the Kozeny-Carman equation is effectively accurate in predicting the hydraulic conductivity of sand;however,when the centrifugal acceleration exceeds 50g,it is important to consider a significant error.
文摘Hydrodynamic cavitation,as an efficient technique applied in many physical and chemical treatment methods,has been widely used by various industries and in several technological fields.Relevant generators,designed with specific structures and parameters,can produce cavitation effects,thereby enabling effective treatment and reasonable transformation of substances.This paper reviews the design principles,performance,and practical applications associated with different types of cavitation generators,aiming to provide theoretical support for the optimization of these systems.It systematically analyzes the underpinning mechanisms and the various factors influencing the cavitation phenomena,also conducting a comparative analysis of the performance of different types of generators.Specific applications dealing with wastewater treatment,chemical reaction acceleration,and other fields are discussed together with the advantages,disadvantages,and applicability of each type of cavitation generator.We also explore research progress in areas such as cavitation stability,energy efficiency,and equipment design upgrades.The study concludes by forecasting the application prospects of intelligent design and computational fluid dynamics(CFD)in optimizing and advancing cavitation generators.It proposes new ideas for the further development of cavitation technology and highlights directions for its widespread future application.
基金the support from the National Natural Science Foundation of China(22478429)the Special Project Fund of Taishan-Scholars(tsqn202408101)+3 种基金the Natural Science Foundation of Shandong Province(ZR2023YQ009)CNPC Innovation Found(2024DQ02-0504)Fundamental Research Funds for the Central Universities,Ocean University of China(202364004)the State Key Laboratory of Heavy Oil Processing(SKLHOP202403003)。
文摘Alcohol oxidation is a widely used green chemical reaction.The reaction process produces flammable and explosive hydrogen,so the design of the reactor must meet stringent safety requirements.Based on the limited experimental data,utilizing the traditional numerical method of computational fluid dynamics(CFD)to simulate the gas-liquid two-phase flow reactor can mitigate the risk of danger under varying working conditions.However,the calculation process is highly time-consuming.Therefore,by integrating process simulation,computational fluid dynamics,and deep learning technologies,an intelligent hybrid chemical model based on machine learning was proposed to expedite CFD calculations,enhance the prediction of flow fields,conversion rates,and concentrations inside the reactor,and offer insights for designing and optimizing the reactor for the alcohol oxidation system.The results show that the hybrid model based on the long and short-term memory neural network achieves 99.8%accuracy in conversion rate prediction and 99.9%accuracy in product concentration prediction.Through validation,the hybrid model is accelerated by about 360 times compared with instrumental analysis in conversion rate prediction and about 45 times compared with CFD calculation in concentration distribution prediction.This hybrid model can quickly predict the conversion rate and product concentration distribution in the gas-liquid two-phase flow reactor and provide a model reference for fast prediction and accurate control in the actual chemical production process.
基金support provided by the National Natural Science Foundation of China(Grant No.12102405)the Presidential Foundation of CAEP(Grant No.YZJJZQ2023008).
文摘This study investigates the paradoxical detonation behavior of TKX-50,a nitrogen-rich energetic material,exhibiting higher detonation velocities but lower metal acceleration ability compared to HMX.Through experimental measurements and theoretical calculations,we propose a novel three-factor competition mechanism to explain this phenomenon.TKX-50-based PBX formulations achieved detonation velocities up to 9100 m/s,surpassing HMX-based counterparts.However,cylinder expansion tests revealed a 15%reduction in metal acceleration ability.Thermochemical measurements showed lower detonation heat for TKX-50(4900 J/g)versus HMX(5645 J/g).Our mechanism involves:(1)compositional effects prevailing at high pressures;(2)Energy release becoming essential as pressure drops;(3)Pressure-dependent product composition evolution functioning at low pressure.VLW code calculations unveiled a"crossover"in Hugoniot curves,lending support to this mechanism.This study furnishes a new framework for comprehending the performance of nitrogen-rich energetic materials,with significant implications for the design and optimization of future high-energy density materials.
基金National Natural Science Foundation of China under Grant No.52278135。
文摘Both acceleration and pseudo-acceleration response spectra play important roles in structural seismic design.However,only one of them is generally provided in most seismic codes.Therefore,many studies have attempted to develop conversion models between the acceleration response spectrum(SA)and the pseudo-acceleration response spectrum(PSA).Our previous studies found that the relationship between SA and PSA is affected by magnitude,distance,and site class.Subsequently,we developed an SA/PSA model incorporating these effects.However,this model is suitable for cases with small and moderate magnitudes and its accuracy is not good enough for cases with large magnitudes.This paper aims to develop an efficient SA/PSA model by considering influences of magnitude,distance,and site class,which can be applied to cases not only with small or moderate magnitudes but also with large ones.For this purpose,regression analyses were conducted using 16,660 horizontal seismic records with a wider range of magnitude.The magnitude of these seismic records varies from 4 to 9 and the distances vary from 10 to 200 km.These ground motions were recorded at 338 stations covering four site classes.By comparing them with existing models,it was found that the proposed model shows better accuracy for cases with any magnitudes,distances,and site classes considered in this study.
基金Supported by the National Natural Science Foundation of China(No.62172352,62171143,42306218)the Guangdong Provincial Department of Education Ocean Ranch Equipment Information and Intelligent Innovation Team Project(No.2023KCXTD016)+1 种基金the Natural Science Foundation of Hebei Province(No.2023407003)the Guangdong Ocean University Research Fund Project(No.060302102304).
文摘In order to solve the problems of limited performance of single and homogeneous sensors and high false alarm rate caused by environmental noise,heterogeneous sensor data fusion(HSDF),a fall detection framework based on heterogeneous sensor fusion of acceleration and audio signals,is proposed in this paper.By analyzing the heterogeneity of acceleration data and audio data,the framework uses Dempstershafer Theory(D-S Theory)to integrate the output of acceleration and audio data at decision level.Firstly,a normalized window interception algorithm——anomaly location window algorithm(ALW)is proposed by analyzing the fall process and the characteristics of acceleration changes based on acceleration data.Secondly,the one-dimensional residual convolutional network(1D-ReCNN)is designed for fall detection based on the audio data.Finally,it is verified that the HSDF framework has good advantages in terms of sensitivity and false alarm rate by the collection of volunteers’simulated fall data and free living data in real environment.
基金supported by the National Natural Science Foundation of China(No.62376198)the National Key Research and Development Program of China(No.2022YFB3104700)the Shanghai Baiyulan Pujiang Project(No.08002360429).
文摘Early exiting has shown significant potential in accelerating the inference of pre-trained language models(PLMs)by allowing easy samples to exit from shallow layers.However,existing early exiting methods primarily rely on local information from individual samples to estimate prediction uncertainty for making exiting decisions,overlooking the global information provided by the sample population.This impacts the estimation of prediction uncertainty,compromising the reliability of exiting de-cisions.To remedy this,inspired by principal component analysis(PCA),the authors define a residual score to capture the deviation of features from the principal space of the sample population,providing a global perspective for estimating prediction uncertainty.Building on this,a two-stage exiting strategy is proposed that integrates global information from residual scores with local information from energy scores at both the decision and feature levels.This strategy incorporates three-way decisions to enable more reliable exiting decisions for boundary region samples by delaying judgement.Extensive experiments on the GLUE benchmark validate that the method achieves an average speed-up ratio of 2.17×across all tasks with minimal per-formance degradation.Additionally,it surpasses the state-of-the-art E-LANG by 11%in model acceleration,along with a performance improvement of 0.6 points,demonstrating a better performance-efficiency trade-off.
基金supported by the National Natural Science Foundation of China (Grant No.52108361)the Sichuan Science and Technology Program of China (Grant No.2023YFS0436)the State Key Laboratory of Geohazard Prevention and Geoenvironment Protection Independent Research Project (Grant No.SKLGP2022Z015).
文摘Bedding parallel stepped rock slopes exist widely in nature and are used in slope engineering.They are characterized by complex topography and geological structure and are vulnerable to shattering under strong earthquakes.However,no previous studies have assessed the mechanisms underlying seismic failure in rock slopes.In this study,large-scale shaking table tests and numerical simulations were conducted to delineate the seismic failure mechanism in terms of acceleration,displacement,and earth pressure responses combined with shattering failure phenomena.The results reveal that acceleration response mutations usually occur within weak interlayers owing to their inferior performance,and these mutations may transform into potential sliding surfaces,thereby intensifying the nonlinear seismic response characteristics.Cumulative permanent displacements at the internal corners of the berms can induce quasi-rigid displacements at the external corners,leading to greater permanent displacements at the internal corners.Therefore,the internal corners are identified as the most susceptible parts of the slope.In addition,the concept of baseline offset was utilized to explain the mechanism of earth pressure responses,and the result indicates that residual earth pressures at the internal corners play a dominant role in causing deformation or shattering damage.Four evolutionary deformation phases characterize the processes of seismic responses and shattering failure of the bedding parallel stepped rock slope,i.e.the formation of tensile cracks at the internal corners of the berm,expansion of tensile cracks and bedding surface dislocation,development of vertical tensile cracks at the rear edge,and rock mass slipping leading to slope instability.Overall,this study provides a scientific basis for the seismic design of engineering slopes and offers valuable insights for further studies on preventing seismic disasters in bedding parallel stepped rock slopes.
基金supported by grants from the National Key Research and Development Program(Grant Nos.2023YFC3806205).
文摘Purpose–This paper aims to analyze the transverse vibration characteristics of the high speed train window glass when passing through tunnel.Design/methodology/approach–The lateral vibration acceleration response of glass chamber of high-speed train CR400BF-A on Beijing-Chengdu high-speed railway was tested at different speeds through the tunnel entrance,exit,tunnel interior,Tunnel Group and rendezvous time in the tunnel,the lateral distribution characteristics of vibration frequency and vibration power amplification coefficient of glass of high-speed train were analyzed.Findings–The results show that:The vibration of the high-speed train glass increases significantly during the tunnel,and the amplitude of vibration acceleration in the tunnel is significantly higher than outside the tunnel as the travel speed increases;the amplitude of lateral vibration acceleration of the glass of a high-speed train does not vary with changes in tunnel length and is not affected by the aerodynamic effects of the tunnel when traveling inside the tunnel,but its vibrations create noticeable fluctuations during variations when encountering oncoming traffic;The vibration characteristics of the high-speed train glass are forced harmonic vibrations,the excitation frequency does not vary with travel speed and travel position changes inside and outside the tunnel.The lateral vibration acceleration of the glass of a high-speed train is applied vertically and uniformly to the glass surface as an“inertial force”and creates a cyclic bending vibration stress that can easily lead to fatigue damage.Originality/value–The research results provide guidance for the prevention of glass failure in high-speed trains.
基金supported in part by the National Natural Science Foundation of China (62422405, 62025111,62495100, 92464302)the STI 2030-Major Projects(2021ZD0201200)+1 种基金the Shanghai Municipal Science and Technology Major Projectthe Beijing Advanced Innovation Center for Integrated Circuits
文摘Computing-in-memory(CIM)has been a promising candidate for artificial-intelligent applications thanks to the absence of data transfer between computation and storage blocks.Resistive random access memory(RRAM)based CIM has the advantage of high computing density,non-volatility as well as high energy efficiency.However,previous CIM research has predominantly focused on realizing high energy efficiency and high area efficiency for inference,while little attention has been devoted to addressing the challenges of on-chip programming speed,power consumption,and accuracy.In this paper,a fabri-cated 28 nm 576K RRAM-based CIM macro featuring optimized on-chip programming schemes is proposed to address the issues mentioned above.Different strategies of mapping weights to RRAM arrays are compared,and a novel direct-current ADC design is designed for both programming and inference stages.Utilizing the optimized hybrid programming scheme,4.67×programming speed,0.15×power saving and 4.31×compact weight distribution are realized.Besides,this macro achieves a normalized area efficiency of 2.82 TOPS/mm2 and a normalized energy efficiency of 35.6 TOPS/W.
基金Science and Technology Plan Project of Xizang Autonomous Region,China under Grant No.XZ202501YD0007。
文摘A novel porous shock absorption layer is put forward in this study, and the shock absorption performance of the porous shock absorption layer is evaluated based on three-dimensional pseudo-static analysis. The modified reaction acceleration method is adopted and validated in the three-dimensional model. Seven ground motions are selected and the peak ground acceleration is adjusted to 0.2 g, 0.4 g and 0.6 g. The impact of the void ratio and thickness of the porous shock absorption layer is studied, while the surrounding rock grade and tunnel depth are also investigated. The numerical results show that the porous shock absorption layer has good shock absorption performance and can effectively reduce the maximum internal force of the secondary lining, but it cannot reduce the maximum horizontal relative displacement of the secondary lining. The circumferential rubber strip in the porous shock absorption layer will reduce shock absorption performance. The results of parameter analysis indicate that the shock absorption performance of the porous shock absorption layer increases with the increase of the void ratio and thickness, and it has good shock absorption performance under different surrounding rock grades and tunnel depths.
基金supported by the Basic Research on Dynamic Real-time Modeling and Onboard Adaptive Modeling of Aero Engine,China(No.QZPY202308)。
文摘Variable Cycle Engine(VCE)serves as the core system in achieving future advanced fighters with cross-generational performance and mission versatility.However,the resultant complex configuration and strong coupling of control parameters present significant challenges in designing acceleration and deceleration control schedules.To thoroughly explore the performance potential of engine,a global integration design method for acceleration and deceleration control schedule based on inner and outer loop optimization is proposed.The outer loop optimization module employs Integrated Surrogate-Assisted Co-Differential Evolutionary(ISACDE)algorithm to optimize the variable geometry adjustment laws based on B-spline curve,and the inner loop optimization module adopts the fixed-state method to design the open-loop fuel–air ratio control schedules,which are aimed at minimizing the acceleration and deceleration time under multiple constraints.Simulation results demonstrate that the proposed global integration design method not only furthest shortens the acceleration and deceleration time,but also effectively safeguards the engine from overlimit.
基金supported by the Fundamental Research Funds for the Central Universities(Grant No.2022CDJQY-012)the Innovation Group Science Foundation of the Natural Science Foundation of Chongqing,China(Grant No.cstc2020jcyj-cxttX0003).
文摘Coralline soils,specialized materials found extensively in the South China Sea,are playing an increasingly vital role in engineering projects.However,like most terrigenous soils,fine-grained coral soil is prone to shrinkage and cracking,which can significantly affect its engineering properties and ultimately jeopardize engineering safety.This paper presents a desiccation cracking test of fine-grained coral soil,with a particular focus on the thickness effect.The study involved measuring the water content and recording the evolution of desiccation cracking.Advanced image processing technology is employed to analyze the variations in crack parameters,clod parameters,fractal dimensions,frequency distributions,and desiccation cracking propagation velocities of fine-grained coral soil.Furthermore,the dynamic evolution of desiccation cracking under the influence of layer thickness is analyzed.A comprehensive crack evolution model is proposed,encompassing both top-down and bottom-up crack propagation,as well as internal tensile cracking.This work introduces novel metrics for the propagation velocity of the total crack area,the characteristic propagation velocities of desiccation cracks,and the acceleration of crack propagation.Through data fitting,theoretical formulas for soil water evaporation,propagation velocities of desiccation cracks,and crack propagation acceleration are derived,laying a foundation for future soil cracking theories.
基金sponsored by National Natural Science Foundation of China (Grant No. 12405215)
文摘This paper introduced a compact high flux polarized neutron beam generator scheme,which used air as the working medium and had low energy consumption.The neutron beam generator adopted a linear three compartment configuration,sequentially using nitrogen nucleus tandem near range accelerated polarization target spallation nuclear reaction technology,neutron multiplication technology,neutron beam polarization and near range acceleration technology,neutron focusing and shooting control technology.Through design and equivalent verification,it has been proven that the total length of the device does not exceed 5 m,the effective range can reach several hundred kilometers,the neutron flux at the muzzle is not less than 10^(25) n·cm^(-2)·s^(-1),which attenuates to 10^(10) n·cm^(-2)·s^(-1) at a distance of several 100 km,and this flux can effectively strike the target.It can be used as a defensive directed energy weapon with high energy density and has broad application prospects.
基金National Natural Science Foundation of China(Grant Nos.52408314,52278292)Chongqing Outstanding Youth Science Foundation(Grant No.CSTB2023NSCQ-JQX0029)+1 种基金Science and Technology Project of Sichuan Provincial Transportation Department(Grant No.2023-ZL-03)Science and Technology Project of Guizhou Provincial Transportation Department(Grant No.2024-122-018).
文摘Lost acceleration response reconstruction is crucial for assessing structural conditions in structural health monitoring(SHM).However,traditional methods struggle to address the reconstruction of acceleration responses with complex features,resulting in a lower reconstruction accuracy.This paper addresses this challenge by leveraging the advanced feature extraction and learning capabilities of fully convolutional networks(FCN)to achieve precise reconstruction of acceleration responses.In the designed network architecture,the incorporation of skip connections preserves low-level details of the network,greatly facilitating the flow of information and improving training efficiency and accuracy.Dropout techniques are employed to reduce computational load and enhance feature extraction.The proposed FCN model automatically extracts high-level features from the input data and establishes a nonlinearmapping relationship between the input and output responses.Finally,the accuracy of the FCN for structural response reconstructionwas evaluated using acceleration data from an experimental arch rib and comparedwith several traditional methods.Additionally,this approach was applied to reconstruct actual acceleration responses measured by an SHM system on a long-span bridge.Through parameter analysis,the feasibility and accuracy of aspects such as available response positions,the number of available channels,and multi-channel response reconstruction were explored.The results indicate that this method exhibits high-precision response reconstruction capability in both time and frequency domains.,with performance surpassing that of other networks,confirming its effectiveness in reconstructing responses under various sensor data loss scenarios.
