Drone swarm systems,equipped with photoelectric imaging and intelligent target perception,are essential for reconnaissance and strike missions in complex and high-risk environments.They excel in information sharing,an...Drone swarm systems,equipped with photoelectric imaging and intelligent target perception,are essential for reconnaissance and strike missions in complex and high-risk environments.They excel in information sharing,anti-jamming capabilities,and combat performance,making them critical for future warfare.However,varied perspectives in collaborative combat scenarios pose challenges to object detection,hindering traditional detection algorithms and reducing accuracy.Limited angle-prior data and sparse samples further complicate detection.This paper presents the Multi-View Collaborative Detection System,which tackles the challenges of multi-view object detection in collaborative combat scenarios.The system is designed to enhance multi-view image generation and detection algorithms,thereby improving the accuracy and efficiency of object detection across varying perspectives.First,an observation model for three-dimensional targets through line-of-sight angle transformation is constructed,and a multi-view image generation algorithm based on the Pix2Pix network is designed.For object detection,YOLOX is utilized,and a deep feature extraction network,BA-RepCSPDarknet,is developed to address challenges related to small target scale and feature extraction challenges.Additionally,a feature fusion network NS-PAFPN is developed to mitigate the issue of deep feature map information loss in UAV images.A visual attention module(BAM)is employed to manage appearance differences under varying angles,while a feature mapping module(DFM)prevents fine-grained feature loss.These advancements lead to the development of BA-YOLOX,a multi-view object detection network model suitable for drone platforms,enhancing accuracy and effectively targeting small objects.展开更多
To maintain the stability of the inter-satellite link for gravitational wave detection,an intelligent learning monitoring and fast warning method of the inter-satellite link control system failure is proposed.Differen...To maintain the stability of the inter-satellite link for gravitational wave detection,an intelligent learning monitoring and fast warning method of the inter-satellite link control system failure is proposed.Different from the traditional fault diagnosis optimization algorithms,the fault intelligent learning method pro-posed in this paper is able to quickly identify the faults of inter-satellite link control system despite the existence of strong cou-pling nonlinearity.By constructing a two-layer learning network,the method enables efficient joint diagnosis of fault areas and fault parameters.The simulation results show that the average identification time of the system fault area and fault parameters is 0.27 s,and the fault diagnosis efficiency is improved by 99.8%compared with the traditional algorithm.展开更多
为探索GS1 Digital Link技术在产品物流中的应用潜力,分析研究了GS1系统和GS1 Digital Link的基本结构、编码特点以及技术优势,充分利用GS1 Digital Link技术可以为产品从源头到零售整个物流过程提供相关对象的Web地址编码的特点,以鲜...为探索GS1 Digital Link技术在产品物流中的应用潜力,分析研究了GS1系统和GS1 Digital Link的基本结构、编码特点以及技术优势,充分利用GS1 Digital Link技术可以为产品从源头到零售整个物流过程提供相关对象的Web地址编码的特点,以鲜活大闸蟹物流过程为例,构建了基于GS1 Digital Link的鲜活大闸蟹Web编码,为实现产品营销与追溯提供了标准化、动态化、多样化的编码数据支撑。展开更多
Von Willebrand factor(VWF)is a multimeric plasma glycoprotein that captures platelets to the sites of vascular injury.The adhesive activity of VWF is highly dependent on the size of VWF.Hemodynamic force converts coil...Von Willebrand factor(VWF)is a multimeric plasma glycoprotein that captures platelets to the sites of vascular injury.The adhesive activity of VWF is highly dependent on the size of VWF.Hemodynamic force converts coiled VWF to linear form,exposing the scissile bond Y1605-M1606 within A2 domain.ADAMTS13(A Disintegrin and Metalloprotease with a ThromboSpondin type 1 motif 13)inhibits excessive platelet aggregation by specifically cleaving the cryptic peptide bond of VWF to regulate its size.Deficiency of ADAMTS13 activity,caused by either mutations or by inhibitory autoantibody,results in the accumulation of ultra-large VWF in plasma,leading to excessive platelet aggregation and disseminated VWF/plateletrich thrombus formation,which is the characteristic of thrombotic thrombocytopenic purpura(TTP).Previous studies showed that,circulating ADAMTS13 generally adopts a closed conformation by the interaction of its TSP8-CUB domain and Spacer domain.This auto-inhibition is relieved when VWF D4-CK domain binds to C-termini of ADAMTS13 disrupting the interaction between TSP8-CUB domain and Spacer domain.As a result,ADAMTS13 changes into an open conformation,exposing more VWF binding sites.Open conformation of ADAMTS13 is considered as a hallmark of acute acquired TTP.However,the dynamic process of conformational transition of ADAMTS13 has not been fully understood.Besides,there are variable natural truncations of ADAMTS13 in circulation,including the truncations lacking the metalloprotease domain.The role of these truncations without enzyme activity in hemostasis is still unknown.Given that there are multiple binding sites in both VWF and open ADAMTS13,we hypothesize that open ADAMTS13 promotes the cross-linking of VWF.Atomic force microscopy(AFM)was employed to prove our hypothesis at single molecule level.The N-termini of ADAMTS13 was immobilized on the polystyrene surface,while the C-termini was stretched by AFM tip.The force required for unfolding was measured and the molecular length increment was obtained by fitting the data with worm-like chain model.In addition,both the polystyrene surface and cantilever were coated with VWF,the adhesion frequency of VWF-VWF interaction were measured in the presence or absence of soluble DisC(a ADAMTS13 truncation that lacks the metalloprotease domain).Our results show that,the rupture force required for ADAMTS13 unfolding is^22 pN.The length increment is mainly in the range of 0-50 nm,and the peak value is 22.6±1.8 nm,consistent with the predicted value of homologous modeling(~27 nm)in literature.The adhesion frequency of two VWF molecules increased in the presence of DisC in a concentration-dependent manner.With or without DisC in solution,the rupture force of the VWF molecules was^20 pN.The bond lifetime of two VWF declined with force increasing,the characteristic of a'slip'bond.In contrast,in the presence of DisC,the bond lifetime firstly increased as force increasing and then decreased as force increasing further,which is the characteristic of a'catch-slip'bond.The distinct patterns of bond lifetime vs force in the presence and absence of DisC indicate that DisC promotes the cross-linking of two VWF molecules.Our data suggest that open ADAMTS13 stabilizes the network of VWF multimers and promotes platelet adhesion.展开更多
社交网络中,节点间存在多种关系类型,节点数量会随着时间的推移而变化,这种异质性和动态性给链路预测任务带来极大的挑战。因此,本文提出一种基于增量学习的社交网络链路预测方法(incremental learning social networks link prediction...社交网络中,节点间存在多种关系类型,节点数量会随着时间的推移而变化,这种异质性和动态性给链路预测任务带来极大的挑战。因此,本文提出一种基于增量学习的社交网络链路预测方法(incremental learning social networks link prediction,IL-SNLP)。通过对网络进行分层,使每一层网络只包含一种关系类型,以更好地获取节点在每种关系类型下的语义信息;针对网络的动态性,利用时序随机游走捕获社交网络中的局部结构信息和时序信息;针对增量数据,采用增量式更新随机游走策略对历史随机游走序列进行更新。通过增量式skip-gram模型从随机游走序列中提取新出现节点的特征,并进一步更新历史节点的特征;针对网络的异质性,采用概率模型提取不同关系类型之间的因果关系关联程度,并将其作用于每一层的节点特征,以改善不同关系层下节点特征表现能力;利用多层感知机构建节点相互感知器,挖掘节点间建立连接时的相互贡献,实现更高的链路预测准确率。实验结果表明,在3个真实的社交网络数据集上,IL-SNLP方法的ROC曲线下的面积(AUC)和F1分数比基线方法分别提高了10.08%~67.60%和1.76%~64.67%,提升了预测性能;对于增量数据,只需要少次迭代就能保持预测模型的性能,提高了模型训练的速度;与未采用增量学习技术的IL-SNLP−方法相比,IL-SNLP方法在时间效率上提升了30.78%~257.58%,显著缩短了模型的运行时长。展开更多
基金supported by the Natural Science Foundation of China,Grant No.62103052.
文摘Drone swarm systems,equipped with photoelectric imaging and intelligent target perception,are essential for reconnaissance and strike missions in complex and high-risk environments.They excel in information sharing,anti-jamming capabilities,and combat performance,making them critical for future warfare.However,varied perspectives in collaborative combat scenarios pose challenges to object detection,hindering traditional detection algorithms and reducing accuracy.Limited angle-prior data and sparse samples further complicate detection.This paper presents the Multi-View Collaborative Detection System,which tackles the challenges of multi-view object detection in collaborative combat scenarios.The system is designed to enhance multi-view image generation and detection algorithms,thereby improving the accuracy and efficiency of object detection across varying perspectives.First,an observation model for three-dimensional targets through line-of-sight angle transformation is constructed,and a multi-view image generation algorithm based on the Pix2Pix network is designed.For object detection,YOLOX is utilized,and a deep feature extraction network,BA-RepCSPDarknet,is developed to address challenges related to small target scale and feature extraction challenges.Additionally,a feature fusion network NS-PAFPN is developed to mitigate the issue of deep feature map information loss in UAV images.A visual attention module(BAM)is employed to manage appearance differences under varying angles,while a feature mapping module(DFM)prevents fine-grained feature loss.These advancements lead to the development of BA-YOLOX,a multi-view object detection network model suitable for drone platforms,enhancing accuracy and effectively targeting small objects.
基金This work was supported by the National Key Research and Development Program Topics(2020YFC2200902)the National Natural Science Foundation of China(11872110).
文摘To maintain the stability of the inter-satellite link for gravitational wave detection,an intelligent learning monitoring and fast warning method of the inter-satellite link control system failure is proposed.Different from the traditional fault diagnosis optimization algorithms,the fault intelligent learning method pro-posed in this paper is able to quickly identify the faults of inter-satellite link control system despite the existence of strong cou-pling nonlinearity.By constructing a two-layer learning network,the method enables efficient joint diagnosis of fault areas and fault parameters.The simulation results show that the average identification time of the system fault area and fault parameters is 0.27 s,and the fault diagnosis efficiency is improved by 99.8%compared with the traditional algorithm.
文摘为探索GS1 Digital Link技术在产品物流中的应用潜力,分析研究了GS1系统和GS1 Digital Link的基本结构、编码特点以及技术优势,充分利用GS1 Digital Link技术可以为产品从源头到零售整个物流过程提供相关对象的Web地址编码的特点,以鲜活大闸蟹物流过程为例,构建了基于GS1 Digital Link的鲜活大闸蟹Web编码,为实现产品营销与追溯提供了标准化、动态化、多样化的编码数据支撑。
基金supported by National Natural Science Foundation of China ( 31500759,31771012)the Guangzhou Science Technology Program ( 201707010062)
文摘Von Willebrand factor(VWF)is a multimeric plasma glycoprotein that captures platelets to the sites of vascular injury.The adhesive activity of VWF is highly dependent on the size of VWF.Hemodynamic force converts coiled VWF to linear form,exposing the scissile bond Y1605-M1606 within A2 domain.ADAMTS13(A Disintegrin and Metalloprotease with a ThromboSpondin type 1 motif 13)inhibits excessive platelet aggregation by specifically cleaving the cryptic peptide bond of VWF to regulate its size.Deficiency of ADAMTS13 activity,caused by either mutations or by inhibitory autoantibody,results in the accumulation of ultra-large VWF in plasma,leading to excessive platelet aggregation and disseminated VWF/plateletrich thrombus formation,which is the characteristic of thrombotic thrombocytopenic purpura(TTP).Previous studies showed that,circulating ADAMTS13 generally adopts a closed conformation by the interaction of its TSP8-CUB domain and Spacer domain.This auto-inhibition is relieved when VWF D4-CK domain binds to C-termini of ADAMTS13 disrupting the interaction between TSP8-CUB domain and Spacer domain.As a result,ADAMTS13 changes into an open conformation,exposing more VWF binding sites.Open conformation of ADAMTS13 is considered as a hallmark of acute acquired TTP.However,the dynamic process of conformational transition of ADAMTS13 has not been fully understood.Besides,there are variable natural truncations of ADAMTS13 in circulation,including the truncations lacking the metalloprotease domain.The role of these truncations without enzyme activity in hemostasis is still unknown.Given that there are multiple binding sites in both VWF and open ADAMTS13,we hypothesize that open ADAMTS13 promotes the cross-linking of VWF.Atomic force microscopy(AFM)was employed to prove our hypothesis at single molecule level.The N-termini of ADAMTS13 was immobilized on the polystyrene surface,while the C-termini was stretched by AFM tip.The force required for unfolding was measured and the molecular length increment was obtained by fitting the data with worm-like chain model.In addition,both the polystyrene surface and cantilever were coated with VWF,the adhesion frequency of VWF-VWF interaction were measured in the presence or absence of soluble DisC(a ADAMTS13 truncation that lacks the metalloprotease domain).Our results show that,the rupture force required for ADAMTS13 unfolding is^22 pN.The length increment is mainly in the range of 0-50 nm,and the peak value is 22.6±1.8 nm,consistent with the predicted value of homologous modeling(~27 nm)in literature.The adhesion frequency of two VWF molecules increased in the presence of DisC in a concentration-dependent manner.With or without DisC in solution,the rupture force of the VWF molecules was^20 pN.The bond lifetime of two VWF declined with force increasing,the characteristic of a'slip'bond.In contrast,in the presence of DisC,the bond lifetime firstly increased as force increasing and then decreased as force increasing further,which is the characteristic of a'catch-slip'bond.The distinct patterns of bond lifetime vs force in the presence and absence of DisC indicate that DisC promotes the cross-linking of two VWF molecules.Our data suggest that open ADAMTS13 stabilizes the network of VWF multimers and promotes platelet adhesion.
文摘社交网络中,节点间存在多种关系类型,节点数量会随着时间的推移而变化,这种异质性和动态性给链路预测任务带来极大的挑战。因此,本文提出一种基于增量学习的社交网络链路预测方法(incremental learning social networks link prediction,IL-SNLP)。通过对网络进行分层,使每一层网络只包含一种关系类型,以更好地获取节点在每种关系类型下的语义信息;针对网络的动态性,利用时序随机游走捕获社交网络中的局部结构信息和时序信息;针对增量数据,采用增量式更新随机游走策略对历史随机游走序列进行更新。通过增量式skip-gram模型从随机游走序列中提取新出现节点的特征,并进一步更新历史节点的特征;针对网络的异质性,采用概率模型提取不同关系类型之间的因果关系关联程度,并将其作用于每一层的节点特征,以改善不同关系层下节点特征表现能力;利用多层感知机构建节点相互感知器,挖掘节点间建立连接时的相互贡献,实现更高的链路预测准确率。实验结果表明,在3个真实的社交网络数据集上,IL-SNLP方法的ROC曲线下的面积(AUC)和F1分数比基线方法分别提高了10.08%~67.60%和1.76%~64.67%,提升了预测性能;对于增量数据,只需要少次迭代就能保持预测模型的性能,提高了模型训练的速度;与未采用增量学习技术的IL-SNLP−方法相比,IL-SNLP方法在时间效率上提升了30.78%~257.58%,显著缩短了模型的运行时长。
