To solve the problem that the existing situation awareness research focuses on multi-sensor data fusion,but the expert knowledge is not fully utilized,a heterogeneous informa-tion fusion recognition method based on be...To solve the problem that the existing situation awareness research focuses on multi-sensor data fusion,but the expert knowledge is not fully utilized,a heterogeneous informa-tion fusion recognition method based on belief rule structure is proposed.By defining the continuous probabilistic hesitation fuzzy linguistic term sets(CPHFLTS)and establishing CPHFLTS distance measure,the belief rule base of the relationship between feature space and category space is constructed through information integration,and the evidence reasoning of the input samples is carried out.The experimental results show that the proposed method can make full use of sensor data and expert knowledge for recognition.Compared with the other methods,the proposed method has a higher correct recognition rate under different noise levels.展开更多
Dempster-Shafer evidence theory is broadly employed in the research of multi-source information fusion.Nevertheless,when fusing highly conflicting evidence it may pro-duce counterintuitive outcomes.To address this iss...Dempster-Shafer evidence theory is broadly employed in the research of multi-source information fusion.Nevertheless,when fusing highly conflicting evidence it may pro-duce counterintuitive outcomes.To address this issue,a fusion approach based on a newly defined belief exponential diver-gence and Deng entropy is proposed.First,a belief exponential divergence is proposed as the conflict measurement between evidences.Then,the credibility of each evidence is calculated.Afterwards,the Deng entropy is used to calculate information volume to determine the uncertainty of evidence.Then,the weight of evidence is calculated by integrating the credibility and uncertainty of each evidence.Ultimately,initial evidences are amended and fused using Dempster’s rule of combination.The effectiveness of this approach in addressing the fusion of three typical conflict paradoxes is demonstrated by arithmetic exam-ples.Additionally,the proposed approach is applied to aerial tar-get recognition and iris dataset-based classification to validate its efficacy.Results indicate that the proposed approach can enhance the accuracy of target recognition and effectively address the issue of fusing conflicting evidences.展开更多
面向开源项目推荐开发人员对开源生态建设具有重要意义。区别于传统软件开发,开源领域的开发者、项目、组织及相互关系体现了开放式协作项目的特点,而它们蕴含的语义有助于精准推荐开源项目的开发者。因此,提出一种基于协作贡献网络(CCN...面向开源项目推荐开发人员对开源生态建设具有重要意义。区别于传统软件开发,开源领域的开发者、项目、组织及相互关系体现了开放式协作项目的特点,而它们蕴含的语义有助于精准推荐开源项目的开发者。因此,提出一种基于协作贡献网络(CCN)的开发者推荐(DRCCN)方法。首先,利用开源软件(OSS)开发者、OSS项目、OSS组织之间的贡献关系构建CCN;其次,基于CCN构建一个3层深度的异构GraphSAGE(Graph SAmple and aggreGatE)图神经网络(GNN)模型,预测开发者节点和开源项目节点之间的链接,从而产生相应的嵌入对;最后,根据预测结果,采用K最近邻(KNN)算法完成开发者推荐。在GitHub数据集上训练和测试模型的实验结果表明,相较于序列推荐的对比学习模型CL4SRec(Contrastive Learning for Sequential Recommendation),DRCCN在精确率、召回率和F1值这3个指标上分别提升了约10.7%、2.6%和4.2%。因此,所提模型可以为开源社区项目的开发者推荐提供重要的参考依据。展开更多
可见光热红外(RGB and Thermal infrared,RGBT)跟踪是一种结合了可见光和热红外光两种不同传感器信息的多模态目标跟踪方法 .这种方法旨在克服单一传感器在特定环境下的局限性,通过融合多种传感器的数据来提高目标跟踪的鲁棒性和准确性...可见光热红外(RGB and Thermal infrared,RGBT)跟踪是一种结合了可见光和热红外光两种不同传感器信息的多模态目标跟踪方法 .这种方法旨在克服单一传感器在特定环境下的局限性,通过融合多种传感器的数据来提高目标跟踪的鲁棒性和准确性.然而,在现有的RGBT跟踪算法中,大多将可见光与热红外图像提取的特征直接进行融合,忽略了两种模态间的同质性与异质性.此外,RGBT跟踪还经常受到目标快速运动、尺度变化、光照变化、热交叉和遮挡等多种挑战因素的影响,现有工作往往是通过研究单一结构来同时解决所有问题,但这需要足够复杂的模型和足够多的训练数据.本文提出了一种新的面向不同挑战并结合多模态同异质信息分离与融合的网络,用于RGBT跟踪.在该网络的每层主干中都设计了一个挑战感知模块用于融合每种挑战下来自可见光与热红外两种不同模态的特征,并自适应地聚合所有挑战下的融合特征.此外,还加入了注意力增强模块及多尺度辅助模块对主干网络所提取的特征进行增强.最后根据可见光与热红外的同质性与异质性,分别提取它们的特有特征与共有特征并进行自适应融合.在GTOT、RGBT234和LasHeR数据集上的大量实验表明,与现有RGBT跟踪方法相比,本文提出的跟踪器显示出非常强的竞争力.展开更多
装备体系组合发展规划是一项复杂的系统工程,具有重要的军事意义和研究价值。考虑到武器装备体系的关联性和复杂性,从体系的角度出发,提出一种装备体系组合发展规划方法,为装备发展论证提供思路。首先,基于装备的不同属性以及关联关系,...装备体系组合发展规划是一项复杂的系统工程,具有重要的军事意义和研究价值。考虑到武器装备体系的关联性和复杂性,从体系的角度出发,提出一种装备体系组合发展规划方法,为装备发展论证提供思路。首先,基于装备的不同属性以及关联关系,构建了装备体系异质网络模型并识别提取功能链,为后续装备发展规划建模提供基础;其次,考虑到装备发展的不确定性,建立了多阶段装备发展规划模型;然后,设计一种双重深度Q网络(double deep Q-network,Double DQN)算法来求解模型;最后,以典型装备体系发展规划为例进行演示计算,验证了所提方法的有效性和可行性。展开更多
为了提高自动驾驶汽车环境感知与安全性,构建高精度的环境地图,提出了基于双目视觉与惯性测量单元(Inertial Measurement Unit, IMU)的多源异构信息融合同步定位与地图构建(Simultaneous Localization and Mapping, SLAM)算法。首先,针...为了提高自动驾驶汽车环境感知与安全性,构建高精度的环境地图,提出了基于双目视觉与惯性测量单元(Inertial Measurement Unit, IMU)的多源异构信息融合同步定位与地图构建(Simultaneous Localization and Mapping, SLAM)算法。首先,针对双目视觉与IMU信息融合的问题,采用紧耦合方法,结合双目视觉传感器的深度感知能力和IMU的快速运动捕捉能力,在系统初始化过程中,引入了一次最大后验估计对双目相机与IMU进行处理;然后,在后端优化中,采用基于滑动窗口的非线性优化算法求解最优位姿;最后,通过自动驾驶试验平台搭建了SLAM系统实物验证平台,设计完成了SLAM系统定位试验和相关性能验证试验。结果表明,双目视觉与IMU信息融合的SLAM系统相较于单目视觉惯性融合(VINS-Fusion)算法的定位精度可提升30.34%,在试验和实际场景中均表现出了有效性。设计的多源异构信息融合的SLAM系统能够显著提升定位精度,且在交通安全环境中具有良好的应用前景,对于提高自动驾驶系统的性能和安全性具有重要意义。展开更多
基金This work was supported by the Youth Foundation of National Science Foundation of China(62001503)the Special Fund for Taishan Scholar Project(ts 201712072).
文摘To solve the problem that the existing situation awareness research focuses on multi-sensor data fusion,but the expert knowledge is not fully utilized,a heterogeneous informa-tion fusion recognition method based on belief rule structure is proposed.By defining the continuous probabilistic hesitation fuzzy linguistic term sets(CPHFLTS)and establishing CPHFLTS distance measure,the belief rule base of the relationship between feature space and category space is constructed through information integration,and the evidence reasoning of the input samples is carried out.The experimental results show that the proposed method can make full use of sensor data and expert knowledge for recognition.Compared with the other methods,the proposed method has a higher correct recognition rate under different noise levels.
基金supported by the National Natural Science Foundation of China(61903305,62073267)the Fundamental Research Funds for the Central Universities(HXGJXM202214).
文摘Dempster-Shafer evidence theory is broadly employed in the research of multi-source information fusion.Nevertheless,when fusing highly conflicting evidence it may pro-duce counterintuitive outcomes.To address this issue,a fusion approach based on a newly defined belief exponential diver-gence and Deng entropy is proposed.First,a belief exponential divergence is proposed as the conflict measurement between evidences.Then,the credibility of each evidence is calculated.Afterwards,the Deng entropy is used to calculate information volume to determine the uncertainty of evidence.Then,the weight of evidence is calculated by integrating the credibility and uncertainty of each evidence.Ultimately,initial evidences are amended and fused using Dempster’s rule of combination.The effectiveness of this approach in addressing the fusion of three typical conflict paradoxes is demonstrated by arithmetic exam-ples.Additionally,the proposed approach is applied to aerial tar-get recognition and iris dataset-based classification to validate its efficacy.Results indicate that the proposed approach can enhance the accuracy of target recognition and effectively address the issue of fusing conflicting evidences.
文摘面向开源项目推荐开发人员对开源生态建设具有重要意义。区别于传统软件开发,开源领域的开发者、项目、组织及相互关系体现了开放式协作项目的特点,而它们蕴含的语义有助于精准推荐开源项目的开发者。因此,提出一种基于协作贡献网络(CCN)的开发者推荐(DRCCN)方法。首先,利用开源软件(OSS)开发者、OSS项目、OSS组织之间的贡献关系构建CCN;其次,基于CCN构建一个3层深度的异构GraphSAGE(Graph SAmple and aggreGatE)图神经网络(GNN)模型,预测开发者节点和开源项目节点之间的链接,从而产生相应的嵌入对;最后,根据预测结果,采用K最近邻(KNN)算法完成开发者推荐。在GitHub数据集上训练和测试模型的实验结果表明,相较于序列推荐的对比学习模型CL4SRec(Contrastive Learning for Sequential Recommendation),DRCCN在精确率、召回率和F1值这3个指标上分别提升了约10.7%、2.6%和4.2%。因此,所提模型可以为开源社区项目的开发者推荐提供重要的参考依据。
文摘可见光热红外(RGB and Thermal infrared,RGBT)跟踪是一种结合了可见光和热红外光两种不同传感器信息的多模态目标跟踪方法 .这种方法旨在克服单一传感器在特定环境下的局限性,通过融合多种传感器的数据来提高目标跟踪的鲁棒性和准确性.然而,在现有的RGBT跟踪算法中,大多将可见光与热红外图像提取的特征直接进行融合,忽略了两种模态间的同质性与异质性.此外,RGBT跟踪还经常受到目标快速运动、尺度变化、光照变化、热交叉和遮挡等多种挑战因素的影响,现有工作往往是通过研究单一结构来同时解决所有问题,但这需要足够复杂的模型和足够多的训练数据.本文提出了一种新的面向不同挑战并结合多模态同异质信息分离与融合的网络,用于RGBT跟踪.在该网络的每层主干中都设计了一个挑战感知模块用于融合每种挑战下来自可见光与热红外两种不同模态的特征,并自适应地聚合所有挑战下的融合特征.此外,还加入了注意力增强模块及多尺度辅助模块对主干网络所提取的特征进行增强.最后根据可见光与热红外的同质性与异质性,分别提取它们的特有特征与共有特征并进行自适应融合.在GTOT、RGBT234和LasHeR数据集上的大量实验表明,与现有RGBT跟踪方法相比,本文提出的跟踪器显示出非常强的竞争力.
文摘装备体系组合发展规划是一项复杂的系统工程,具有重要的军事意义和研究价值。考虑到武器装备体系的关联性和复杂性,从体系的角度出发,提出一种装备体系组合发展规划方法,为装备发展论证提供思路。首先,基于装备的不同属性以及关联关系,构建了装备体系异质网络模型并识别提取功能链,为后续装备发展规划建模提供基础;其次,考虑到装备发展的不确定性,建立了多阶段装备发展规划模型;然后,设计一种双重深度Q网络(double deep Q-network,Double DQN)算法来求解模型;最后,以典型装备体系发展规划为例进行演示计算,验证了所提方法的有效性和可行性。
文摘为了提高自动驾驶汽车环境感知与安全性,构建高精度的环境地图,提出了基于双目视觉与惯性测量单元(Inertial Measurement Unit, IMU)的多源异构信息融合同步定位与地图构建(Simultaneous Localization and Mapping, SLAM)算法。首先,针对双目视觉与IMU信息融合的问题,采用紧耦合方法,结合双目视觉传感器的深度感知能力和IMU的快速运动捕捉能力,在系统初始化过程中,引入了一次最大后验估计对双目相机与IMU进行处理;然后,在后端优化中,采用基于滑动窗口的非线性优化算法求解最优位姿;最后,通过自动驾驶试验平台搭建了SLAM系统实物验证平台,设计完成了SLAM系统定位试验和相关性能验证试验。结果表明,双目视觉与IMU信息融合的SLAM系统相较于单目视觉惯性融合(VINS-Fusion)算法的定位精度可提升30.34%,在试验和实际场景中均表现出了有效性。设计的多源异构信息融合的SLAM系统能够显著提升定位精度,且在交通安全环境中具有良好的应用前景,对于提高自动驾驶系统的性能和安全性具有重要意义。
