Complex networks have attracted much attention from various fields of sciences and engineering in recent years.However, many complex networks have various uncertain information, such as unknown or uncertain system par...Complex networks have attracted much attention from various fields of sciences and engineering in recent years.However, many complex networks have various uncertain information, such as unknown or uncertain system parameters and topological structure, which greatly affects the system dynamics.Thus, the parameter estimation and structure identification problem has theoretical and practical importance for uncertain complex dynamical networks. This paper investigates identification of unknown system parameters and network topologies in uncertain fractional-order complex network with time delays(including coupling delay and node delay). Based on the stability theorem of fractional-order differential system and the adaptive control technique, a novel and general method is proposed to address this challenge. Finally two representative examples are given to verify the effectiveness of the proposed approach.展开更多
近年来,利用高阶交互信息进行多层网络社区检测已成为复杂网络分析领域的研究热点。尽管多层网络社区检测的研究已取得了一些进展,但大多数方法忽略了网络各层之间的联系。为了解决这一问题,提出了一种模体(motif)感知的自适应跨层游走...近年来,利用高阶交互信息进行多层网络社区检测已成为复杂网络分析领域的研究热点。尽管多层网络社区检测的研究已取得了一些进展,但大多数方法忽略了网络各层之间的联系。为了解决这一问题,提出了一种模体(motif)感知的自适应跨层游走社区检测算法(Motif-aware Adaptive Cross-Layer random walk Community Detection,MACLCD)。该算法充分考虑了多层网络各层内的高阶交互特性以及层间的相关性,有效整合了多层网络的结构信息,提高了社区检测结果的准确性。具体地,首先从网络和节点的角度进行综合度量,揭示网络层间相关性;其次,考虑了各层网络可能具有不同的局部和全局结构特征,利用motif识别各层网络特有的高阶交互结构,构建多层加权混合阶网络;进一步,设计了多层网络跨层游走模型,并引入跳转因子,以确保随机游走能够自适应地遍历多层网络,从而捕获更丰富的网络结构信息。在4个真实的网络数据集上进行实验比较分析,结果表明MACLCD算法在社区检测方面性能较优,相比目前表现最佳的对比算法在ACC和NMI上分别提高了10%和8.9%。展开更多
随着功率模块集成化程度的提高,其散热结构优化已成为研发中的关键。拓扑优化可通过变换散热器形貌、结构来最大化地提升散热效果,因此受到了广泛关注。但在拓扑优化过程中,每步迭代均需要计算模块与散热器温度分布,占用较庞大的计算资...随着功率模块集成化程度的提高,其散热结构优化已成为研发中的关键。拓扑优化可通过变换散热器形貌、结构来最大化地提升散热效果,因此受到了广泛关注。但在拓扑优化过程中,每步迭代均需要计算模块与散热器温度分布,占用较庞大的计算资源和计算时间。为加速传统散热器拓扑优化进程,在基于传统固体各向同性材料惩罚SIMP(solid isotropic material with penalization)散热器拓扑优化方法的基础上,提出一种嵌套神经网络NN(neural network)同步学习的快速迭代方法。首先,构建散热器基于编码器-解码器结构的NN预测模型,即基于散热器形貌迭代进化过程实现优化结构的快速预测;其次,将NN模型与散热器SIMP拓扑优化流程相嵌套,利用迭代过程中的中间形貌同步训练NN;最后,针对单芯片、两芯片模块结构,对比所提方法与传统迭代方法的拓扑优化结果,验证了所提NN同步学习方法的准确性和快速性。展开更多
基金supported by the Basic and Frontier Research Project of Chongqing(cstc2013jcyj A70006,cstc2015jcyj A40038)
文摘Complex networks have attracted much attention from various fields of sciences and engineering in recent years.However, many complex networks have various uncertain information, such as unknown or uncertain system parameters and topological structure, which greatly affects the system dynamics.Thus, the parameter estimation and structure identification problem has theoretical and practical importance for uncertain complex dynamical networks. This paper investigates identification of unknown system parameters and network topologies in uncertain fractional-order complex network with time delays(including coupling delay and node delay). Based on the stability theorem of fractional-order differential system and the adaptive control technique, a novel and general method is proposed to address this challenge. Finally two representative examples are given to verify the effectiveness of the proposed approach.
文摘近年来,利用高阶交互信息进行多层网络社区检测已成为复杂网络分析领域的研究热点。尽管多层网络社区检测的研究已取得了一些进展,但大多数方法忽略了网络各层之间的联系。为了解决这一问题,提出了一种模体(motif)感知的自适应跨层游走社区检测算法(Motif-aware Adaptive Cross-Layer random walk Community Detection,MACLCD)。该算法充分考虑了多层网络各层内的高阶交互特性以及层间的相关性,有效整合了多层网络的结构信息,提高了社区检测结果的准确性。具体地,首先从网络和节点的角度进行综合度量,揭示网络层间相关性;其次,考虑了各层网络可能具有不同的局部和全局结构特征,利用motif识别各层网络特有的高阶交互结构,构建多层加权混合阶网络;进一步,设计了多层网络跨层游走模型,并引入跳转因子,以确保随机游走能够自适应地遍历多层网络,从而捕获更丰富的网络结构信息。在4个真实的网络数据集上进行实验比较分析,结果表明MACLCD算法在社区检测方面性能较优,相比目前表现最佳的对比算法在ACC和NMI上分别提高了10%和8.9%。
文摘随着功率模块集成化程度的提高,其散热结构优化已成为研发中的关键。拓扑优化可通过变换散热器形貌、结构来最大化地提升散热效果,因此受到了广泛关注。但在拓扑优化过程中,每步迭代均需要计算模块与散热器温度分布,占用较庞大的计算资源和计算时间。为加速传统散热器拓扑优化进程,在基于传统固体各向同性材料惩罚SIMP(solid isotropic material with penalization)散热器拓扑优化方法的基础上,提出一种嵌套神经网络NN(neural network)同步学习的快速迭代方法。首先,构建散热器基于编码器-解码器结构的NN预测模型,即基于散热器形貌迭代进化过程实现优化结构的快速预测;其次,将NN模型与散热器SIMP拓扑优化流程相嵌套,利用迭代过程中的中间形貌同步训练NN;最后,针对单芯片、两芯片模块结构,对比所提方法与传统迭代方法的拓扑优化结果,验证了所提NN同步学习方法的准确性和快速性。
