Based on the knot theory and researching of network structures of glucomannan molecules, the polysaccharides were analyzed. The link prediction analysis is to further reveal the interactions between polysaccharides, t...Based on the knot theory and researching of network structures of glucomannan molecules, the polysaccharides were analyzed. The link prediction analysis is to further reveal the interactions between polysaccharides, to elaborate QSAR of polysaccharides, and to analyze the network conformation relationships among polysaccharides. We made a classification for glucomannan molecules based on the related domestic and international theories, and investigated their network structures and application prospects. The knot theory and the link predictions not only simplify the glucomannan microscopic descriptions but also play a guiding role in predicting and regulating the structures.展开更多
The dynamic changes of the complex network and the material form and function were actuated by the molecular chains. The interaction behavior between molecular chains was difficult to illuminate because the dynamic ch...The dynamic changes of the complex network and the material form and function were actuated by the molecular chains. The interaction behavior between molecular chains was difficult to illuminate because the dynamic changes of macromolecules were observed difficultly by normal spectrum method and the methods to test and evaluate the complex network evolution prediction and intervention are rare. The mathematic model of domino offect of molecular chains was established based on the topological structure of molecular chain aggregation of Konjac glucomannan, and the molecular entanglement mechanism of Konjac glucamannan blends was studied through molecular simulation and knot theory analysis combined with experimental verification. The results suggested that two network models （topological entanglement and solid knot） of Konjac glucomannon blends were formed through hydrogen bond nodes. The topological entanglement was strengthened with the increase of concentration and the form of molecular chains was Gaussian chain which could not allow traverse moving owing to the intermolecular cross and entanglement and the shield of intramolecular interaction. Besides, the structures of Konjac glucomannon blends became more stable due to the solid knot. Both of them were verified by the experimental results. This experimental method simplifies the microscopic description of Konjac glucomannon, and there is important guiding significance of the experimental results for the prediction and control ofpolysaccharides＇ structure and function.展开更多
节子影响着实木板材力学性能,如何准确刻画出节子在实木板材内部的形态,进而计算出实木板材力学性能是一个具有应用价值的科学问题。目前,基于机器视觉的缺陷检测方法实现了实木板材表面缺陷检测与识别,超声波检测方法可以判断出实木板...节子影响着实木板材力学性能,如何准确刻画出节子在实木板材内部的形态,进而计算出实木板材力学性能是一个具有应用价值的科学问题。目前,基于机器视觉的缺陷检测方法实现了实木板材表面缺陷检测与识别,超声波检测方法可以判断出实木板材中缺陷的存在, X-ray虽然可以全面的掌握实木信息,但其检测成本较高。近红外光谱分析具有结构丰富,测试方便、无损快速的特点,但是,光谱中存在的冗余与非线性信息影响建模精准度,提出一种基于Isomap和小波神经网络融合的节子倾角辨识方法,利用Isomap完成光谱信息非线性降维,运用小波神经网络建立节子边缘的物质成分与倾角间的非线性关联,通过边缘多点倾角反演出节子在实木板材内部的形态。首先,采用Pablo提出的节子斜圆锥模型,并结合图像处理提取实木板材表面的节子缺陷区域,计算出相应中心位置;提取节子边缘的多点位置,采集光谱信息并完成基线漂移和去噪处理;然后,利用K-S划分校正样本集,运用主成分与马氏距离结合剔除异常光谱;接着,运用Isomap方法设定降维数和邻近数,通过PLS完成不同光谱维度的快速建模,进而迭代出理想光谱特征;最后,应用具有局部信息优化能力的小波神经网络建立节子边缘光谱与该点倾角间的非线性关系,构建出1个12输入、 1输出的网络模型,并运用梯度修正网络参数;将节子倾角预测结果输入Solidworks软件完成节子椎体形态的三维呈现。实验采用落叶松实木板材作为对象,选取并采集了40个节子的160组光谱数据,通过测量上、下表面节子的相对空间位置,计算出边缘点倾斜角数值并进行建模分析,实验结果表明:采用S-G平滑与一阶导数进行光谱预处理,得到的光谱轮廓更清晰、吸收峰更明显;采用Isomap特征降维方法,选取非线性降维数 d =12、近邻数 k =19时, SECV最小,可以消除光谱信息的冗余数据;采用小波神经网络建立的节子倾角非线性模型,其预测相关系数为0.88,预测标准差为7.65,相对分析误差为2.14;可以实现节子在实木板材内部的形态反演,可以为力学性能预测提供定量化分析手段。展开更多
基金Supported by the National Natural Science Foundation of China(31271837 and 31071518)Specialized Research Fund for the Doctoral Program of Higher Education jointly funded by Ministry of Education(20113515110010)+2 种基金Special Research Funds from Ministry of Science and Technology(2012GA7200022)Major projects of industries,universities and research in Fujian Province(2013N5003)Natural Science Foundation of Fujian Province(2011J0101)
文摘Based on the knot theory and researching of network structures of glucomannan molecules, the polysaccharides were analyzed. The link prediction analysis is to further reveal the interactions between polysaccharides, to elaborate QSAR of polysaccharides, and to analyze the network conformation relationships among polysaccharides. We made a classification for glucomannan molecules based on the related domestic and international theories, and investigated their network structures and application prospects. The knot theory and the link predictions not only simplify the glucomannan microscopic descriptions but also play a guiding role in predicting and regulating the structures.
基金supported by the National Natural Science Foundation of China(31271837)Specialized Research Fund for the Doctoral Program of Higher Education jointly funded by Ministry of Education(20113515110010)+2 种基金Special Research Funds from Ministry of Science and Technology(2012GA7200022)Major projects of industries,universities and research in Fujian Province(2013N5003)Natural Science Foundation of Fujian Province(2011J0101)
文摘The dynamic changes of the complex network and the material form and function were actuated by the molecular chains. The interaction behavior between molecular chains was difficult to illuminate because the dynamic changes of macromolecules were observed difficultly by normal spectrum method and the methods to test and evaluate the complex network evolution prediction and intervention are rare. The mathematic model of domino offect of molecular chains was established based on the topological structure of molecular chain aggregation of Konjac glucomannan, and the molecular entanglement mechanism of Konjac glucamannan blends was studied through molecular simulation and knot theory analysis combined with experimental verification. The results suggested that two network models （topological entanglement and solid knot） of Konjac glucomannon blends were formed through hydrogen bond nodes. The topological entanglement was strengthened with the increase of concentration and the form of molecular chains was Gaussian chain which could not allow traverse moving owing to the intermolecular cross and entanglement and the shield of intramolecular interaction. Besides, the structures of Konjac glucomannon blends became more stable due to the solid knot. Both of them were verified by the experimental results. This experimental method simplifies the microscopic description of Konjac glucomannon, and there is important guiding significance of the experimental results for the prediction and control ofpolysaccharides＇ structure and function.
文摘节子影响着实木板材力学性能,如何准确刻画出节子在实木板材内部的形态,进而计算出实木板材力学性能是一个具有应用价值的科学问题。目前,基于机器视觉的缺陷检测方法实现了实木板材表面缺陷检测与识别,超声波检测方法可以判断出实木板材中缺陷的存在, X-ray虽然可以全面的掌握实木信息,但其检测成本较高。近红外光谱分析具有结构丰富,测试方便、无损快速的特点,但是,光谱中存在的冗余与非线性信息影响建模精准度,提出一种基于Isomap和小波神经网络融合的节子倾角辨识方法,利用Isomap完成光谱信息非线性降维,运用小波神经网络建立节子边缘的物质成分与倾角间的非线性关联,通过边缘多点倾角反演出节子在实木板材内部的形态。首先,采用Pablo提出的节子斜圆锥模型,并结合图像处理提取实木板材表面的节子缺陷区域,计算出相应中心位置;提取节子边缘的多点位置,采集光谱信息并完成基线漂移和去噪处理;然后,利用K-S划分校正样本集,运用主成分与马氏距离结合剔除异常光谱;接着,运用Isomap方法设定降维数和邻近数,通过PLS完成不同光谱维度的快速建模,进而迭代出理想光谱特征;最后,应用具有局部信息优化能力的小波神经网络建立节子边缘光谱与该点倾角间的非线性关系,构建出1个12输入、 1输出的网络模型,并运用梯度修正网络参数;将节子倾角预测结果输入Solidworks软件完成节子椎体形态的三维呈现。实验采用落叶松实木板材作为对象,选取并采集了40个节子的160组光谱数据,通过测量上、下表面节子的相对空间位置,计算出边缘点倾斜角数值并进行建模分析,实验结果表明:采用S-G平滑与一阶导数进行光谱预处理,得到的光谱轮廓更清晰、吸收峰更明显;采用Isomap特征降维方法,选取非线性降维数 d =12、近邻数 k =19时, SECV最小,可以消除光谱信息的冗余数据;采用小波神经网络建立的节子倾角非线性模型,其预测相关系数为0.88,预测标准差为7.65,相对分析误差为2.14;可以实现节子在实木板材内部的形态反演,可以为力学性能预测提供定量化分析手段。
