基于深度学习的ELM实时识别研究

基于深度学习的方法，在HL-2A装置上开发出了一套边缘局域模(ELM)实时识别算法。算法使用5200次放电数据(约24.19万数据切片)进行学习，得到一个深度为22层的卷积神经网络。为衡量算法的识别能力，识别了HL-2A装置自2009年实现稳定ELMy H模放电以来所有历史数据(约26000次放电数据)，共识别出1665次H模放电，其中误识别35次，误报率为2.10%。在实际的1634次H模放电中，漏识别4次，漏识别率为0.24%。该误报率和漏报率可以满足ELM实时识别的精度要求。识别算法在实时控制环境下，对单个时间点的平均计算时间为0.46ms，可以满足实时控制的计算速度要求。     

文化创意企业视角的长三角城市网络演变特征

运用社会网络与位序-规模法分析长江三角洲地区16市上市类文化创意企业的母-子公司数据信息, 揭示文化创意企业集聚视角长三角城市网络结构及其演变特征. 研究发现: (1) 2005~2015年, 长三角地区文化创意企业上市公司总部数量逐年增加, 公司总部多设立于上海、杭州、苏州以及南京4地; (2)传媒类和软件及计算机类企业多分布于传统区域, 设计咨询类企业生产网络中, 存在传统非核心城市跃升为区域核心城市的现象; (3)上市公司生产网络结构逐渐复杂, 多中心及新核心相继出现, 且生产网络中城市位序有变动趋势, 文化创意产业空间动态正形塑城市体系结构. 为此, 建议长三角文化创意企业发展中应加强企业母子间的联系强度, 促进城市创意流动网络化, 有望推进长三角创意源-流形塑创意城市群新体系.     

基于显示连通贝叶斯网络的机场塔台地震易损性分析

将基于性能的多维易损性分析方法,结合显示连通贝叶斯网络,应用于机场塔台的多维易损性分析。考虑地震激励的不确定性,通过非线性时程分析获得结构响应数据;将塔台结构分为三个层次,每个层次按包含的层数分为相应的子层次。根据功能特性确定子层次的评价指标和极限状态,建立服从多元对数正态分布的概率地震需求模型;考虑各种极限状态之间的相关性,建立极限状态方程,确定失效域,通过蒙特卡洛法求得构件的超越概率;建立塔台结构的显示连通贝叶斯网络模型,利用层次分析法获得中间节点的条件概率表,利用MATLAB进行贝叶斯网络的推理计算,实现从单一层次的易损性到整体易损性的推理。     

卷积神经网络用于近红外光谱古筝面板木材分级

目前,我国乐器制作行业在古筝面板用木材等级的筛选上主要依赖于技师主观评判,但此法缺少科学理论的依据,效率低,客观性及出材率的提高等方面受到限制,无法满足乐器市场的大量需求。实现古筝面板用木材快速、智能化的分级工作是一个急需解决的课题。近红外光谱非常适用于测量含氢的有机物质。古筝面板木材主要化学成分的化学键均由含氢基团组成,不同等级板材的化学成分存在差异,这些差异反映在近红外光谱中,为判断木材等级提供了可能。同时卷积神经网络对非线性数据具有较强的特征提取能力,所以提出一种应用卷积神经网络模型对光谱数据进行分析的方法,进而判别木材的等级。应用了Savitzky Golay一阶、二阶微分两种预处理方法和核主成分分析、连续投影算法两种数据压缩方法,通过所设计的卷积神经网络模型以样本识别准确率和模型构建过程中的损失值作为判定指标选出最佳预处理和数据压缩方法。为了提高模型提取分析光谱数据的能力和避免过拟合现象,应用了多通道卷积核、批量归一化和early stopping策略,将通过两层卷积层提取的特征信息送入全连接层,从而充分提取剩余信息,通过Softmax函数获得板材的最终预测等级,从而确定了最终模型。最终Savitzky Golay一阶微分和核主成分分析为最佳数据处理方法,同时得出用于区分不同等级的古筝面板用木材的主要关键谱带,分别为1 163~1 243, 1 346~1 375和1 525~1 584 nm。将该模型应用于测试集样本,古筝面板用木材的等级识别准确率为95.5%。实验结果表明所提出的方法可以高效地处理光谱数据,有效识别区分不同等级的古筝面板用木材的关键特征,从而为广阔的乐器市场提供一定的技术支持。     

Convolutional Neural Network Approach for Multispectral Facial Presentation Attack Detection in Automated Border Control Systems

Automated border control systems are the first critical infrastructure point when crossing a border country. Crossing border lines for unauthorized passengers is a high security risk to any country. This paper presents a multispectral analysis of presentation attack detection for facial biometrics using the learned features from a convolutional neural network. Three sensors are considered to design and develop a new database that is composed of visible (VIS), near-infrared (NIR), and thermal images. Most studies are based on laboratory or ideal conditions-controlled environments. However, in a real scenario, a subject’s situation is completely modified due to diverse physiological conditions, such as stress, temperature changes, sweating, and increased blood pressure. For this reason, the added value of this study is that this database was acquired in situ. The attacks considered were printed, masked, and displayed images. In addition, five classifiers were used to detect the presentation attack. Note that thermal sensors provide better performance than other solutions. The results present better outputs when all sensors are used together, regardless of whether classifier or feature-level fusion is considered. Finally, classifiers such as KNN or SVM show high performance and low computational level.     

Neural network representations of quantum many-body states

Machine learning is currently the most active interdisciplinary field having numerous applications;additionally,machine-learning techniques are used to research quantum many-body problems.In this study,we first propose neural network quantum states(NNQSs)with general input observables and explore a few related properties,such as the tensor product and local unitary operation.Second,we determine the necessary and sufficient conditions for the representability of a general graph state using normalized NNQS.Finally,to quantify the approximation degree of a given pure state,we define the best approximation degree using normalized NNQSs.Furthermore,we observe that some 7V-qubit states can be represented by a normalized NNQS,such as separable pure states,Bell states and GHZ states.     

Design and build an elastic crosslinked network to strengthen and toughen soybean-meal based bioadhesive using organo-sepiolite and greener crosslinker triglycidylamine

Poor water resistance and curing layer brittleness are significant challenges of greener soybean meal (SM) adhesive promotion and application. In this study an elastic network was built to crosslink the SM adhesive and improve the cured layer toughness of the resultant adhesive. Long-chain organo-sepiolite (OSEP) was first prepared by compound modification using KH-560 and KH-602 silane coupling agents. Triglycidylamine (TGA), a greener crosslinking agent with a large number of effective epoxy functional groups, was then synthesized. SM-based adhesives were fabricated using SM, OSEP and TGA. Results indicated that an elastic network was fabricated by the cross-linked reaction of SM, OSEP, and TGA. The elastic network effectively improved the toughness of the resultant adhesive. The wet shear strength of the plywood (PlyW) bonded by SM with 1% OSEP (SM/TGA/OSEP-1 adhesive) increased by 60% to 1.25 MPa relative to that of the SM/TGA adhesive. This study provided a greener, simple and cheap method improving the toughness and wet shear strength of SM-based adhesive, which contributes to the industrialized application of the product.     

Synthesis of Heat-resistant Polymers by Thiol–Ene Reaction of N-Allylmaleimide Copolymers Using Glycoluril Crosslinkers with Rigid Molecular Structures

We carried out the thermal curing of the copolymers of N-allylmaleimide (AMI) and 2-ethylhexyl acrylate (2EHA) using 1,3,4,6-tetra(2-mercaproethyl)glycoluril ( G1 ), 1,3,4,6-tetra(3-mercaptopropyl)glycoluril ( G2 ), 1,3,4,6-tetraallylglycoluril ( G3 ), triallylisocyanurate (TAIC), and pentaerythritol tetrakis(3-mercaptobutyrate) (PEMB) as the crosslinkers. Based on the results for the analysis of thiol–ene reactions monitored by IR spectroscopy, it was confirmed that the curing rate significantly depended on the combination of the used crosslinkers. The insoluble fraction after curing was more than 90% for the systems using the glycoluril crosslinkers, while the conversion of the allyl groups was suppressed due to the rigid structure of these crosslinkers. The heat resistance and the mechanical properties of the crosslinked polymers were investigated by thermogravimetric analysis, differential scanning calorimetry, dynamic mechanical analysis, and mechanical tensile tests. For the products cured using the glycoluril crosslinkers, the glass transition temperature (T_g) and the maximum temperature of thermal decomposition (T_max) were 54–59 °C and 395–409 °C, respectively, being higher than those for the cured product prepared with PEMB and TAIC as the conventional crosslinkers. The elasticity (75–139 MPa), the maximum strength (3.0–4.1 MPa), and the adhesion strength (6.7–10.7 MPa) for the polymers cured with the glycoluril crosslinkers, determined by the mechanical tensile and single lap-shear adhesion tests, were higher than those for cured materials produced with PEMB. Thus, the thermal and mechanical properties of the maleimide copolymers were efficiently enhanced by crosslinking using the rigid glycoluril compounds. © 2020 Wiley Periodicals, Inc. J. Polym. Sci. 2020 , 58, 923–931     

Highly stretchable,self-adhesive,and self-healable double network hydrogel based on alginate/polyacrylamide with tunable mechanical properties

A number of synthetic hydrogels suffer from low mechanical strength. Despite of the recent advances in the fabrication of tough hydrogels, it is still a great challenge to simultaneously construct high stretchability, and self-adhesive and self-healing capability in a hydrogel. Herein, a new type of double network hydrogel was prepared based on irreversible cross-linking of polyacrylamide chains and Schiff-base reversible cross-linking between glycidyl methacrylate-grafted ethylenediamine and oxidized sodium alginate (OSA). The combination of both cross-linkings and their synergistic effect provided a novel hydrogel with high strength, stretchable, rapid self-healing, and self-adhesiveness to different material. Besides, the hydrogels with diverse OSA content could maintain their original shapes after loading–unloading tensile test. The resulting hydrogel has a great potential in various fields for supporting and load-bearing substance.     

A post curing strategy toward the feasible covalent adaptable networks in polyacrylate latex films

The implementation of covalent adaptable networks (CANs) in general resin system is becoming attractive. In this work, we propose a simple post-curing strategy based on the core-shell structured acrylate latex for the achievement on both the improved general performance and the CANs characteristics in latex films. The building to the CANs was relied on the introduction of 4,4′-diaminophenyl disulfide as the curing agent, which cured the acetoacetoxy decorated shell polymer through the ketoamine reaction. The metathesis reaction of aromatic disulfides in the crosslinking segments enabled the thermally induced dynamic behavior of the network as revealed in the stress relaxation tests by comparison with other diamine crosslinking agents without the incorporation of disulfide. The synergism of the dynamic crosslinking of the shell polymer and static crosslinking in the core polymer contributed to the improved mechanical strength (15 MPa, strain% = 250%) and the suppressed water adsorption (~1% in 24 h of soaking) of the latex film, which exhibited above 90% of recovery in both strength and strain from a cut-off film damage within 1 h at 80°C. Moreover, the cured latex film could be recycled, and 75% of the mechanical performance was regained after three fragmentation-hot-pressing cycles. These, in addition with the feasible and environmental friendly characteristics, suggest a sustainable paradigm toward the smart thermosetting latex polymers.