In this paper,a new kind of flexible cone composed of the thin-walled plates based on space probecone docking mechanism for small-sized spacecraft is presented.The theoretical model of docking impact dynamics,which ta...In this paper,a new kind of flexible cone composed of the thin-walled plates based on space probecone docking mechanism for small-sized spacecraft is presented.The theoretical model of docking impact dynamics,which takes into account the additional stiffness terms,is derived based on Lagrange Analytical Mechanics theory and Hertz contact theory.Finite element method is employed for the discretization of the thin-walled plate.The results show that the traditional dynamic model without considering the additional stiffness terms will be difficult to reach steady state.The method proposed in this paper can correctly predict the dynamic behavior of the system.展开更多
在结构健康监测系统中重构缺失响应数据对于准确评估结构工作状况至关重要。提出了一种基于双向长短期记忆网络和注意力机制的缺失振动响应重构网络——序列到序列-双向长短时记忆网络-注意力模型。该网络在序列到序列(sequence to sequ...在结构健康监测系统中重构缺失响应数据对于准确评估结构工作状况至关重要。提出了一种基于双向长短期记忆网络和注意力机制的缺失振动响应重构网络——序列到序列-双向长短时记忆网络-注意力模型。该网络在序列到序列(sequence to sequence,Seq2Seq)架构的基础上,将响应重构问题建模为序列生成问题,利用数据间潜在的时空关系显著提高模型的重构性能。此外,提出了一种基于均值平滑的损失计算方法评估模型的整体性能。通过对八自由度振动系统数值算例以及道林厅人行桥实际监测数据的研究,验证了所提出模型的鲁棒性与准确性。试验结果表明,该模型在不同噪声环境下均能胜任响应重构任务,在低信噪比的情况下仍表现出优异的重构性能。展开更多
基金Supported by the National Natural Science Foundation of China(91216201,11725211)
文摘In this paper,a new kind of flexible cone composed of the thin-walled plates based on space probecone docking mechanism for small-sized spacecraft is presented.The theoretical model of docking impact dynamics,which takes into account the additional stiffness terms,is derived based on Lagrange Analytical Mechanics theory and Hertz contact theory.Finite element method is employed for the discretization of the thin-walled plate.The results show that the traditional dynamic model without considering the additional stiffness terms will be difficult to reach steady state.The method proposed in this paper can correctly predict the dynamic behavior of the system.
文摘在结构健康监测系统中重构缺失响应数据对于准确评估结构工作状况至关重要。提出了一种基于双向长短期记忆网络和注意力机制的缺失振动响应重构网络——序列到序列-双向长短时记忆网络-注意力模型。该网络在序列到序列(sequence to sequence,Seq2Seq)架构的基础上,将响应重构问题建模为序列生成问题,利用数据间潜在的时空关系显著提高模型的重构性能。此外,提出了一种基于均值平滑的损失计算方法评估模型的整体性能。通过对八自由度振动系统数值算例以及道林厅人行桥实际监测数据的研究,验证了所提出模型的鲁棒性与准确性。试验结果表明,该模型在不同噪声环境下均能胜任响应重构任务,在低信噪比的情况下仍表现出优异的重构性能。
