This article presents a novel method to prove that: let E be an AM-space and if dim E ≥ 3, then there does not exist any odd subtractive.isometric mapping from the unit sphere S(E) into S[L(Ω, μ)]. In particul...This article presents a novel method to prove that: let E be an AM-space and if dim E ≥ 3, then there does not exist any odd subtractive.isometric mapping from the unit sphere S(E) into S[L(Ω, μ)]. In particular, there does not exist any real linear isometry from E into L(Ω, μ).展开更多
针对化工连续生产过程的时序性及非线性等特征,提出一种新的基于数据驱动的化工过程故障诊断方法:ISOMAP-LDA。首先实行流形学习算法ISOMAP,在保持量测数据几何结构特性下完成非线性降维,然后基于提取的嵌入变量张成的低维空间,选用线...针对化工连续生产过程的时序性及非线性等特征,提出一种新的基于数据驱动的化工过程故障诊断方法:ISOMAP-LDA。首先实行流形学习算法ISOMAP,在保持量测数据几何结构特性下完成非线性降维,然后基于提取的嵌入变量张成的低维空间,选用线性判别分析(LDA)构造故障模式类的判别函数,负责各采样个体故障类型的判定。将该方法用于仿真化工Tennessee East man过程的故障诊断,结果表明,ISOMAP-LDA方法不仅拥有较高的故障诊断能力,而且取得采样在低维空间的可视化表示。展开更多
基金This study is supported by the National Natural Science Foundation of China (10571090)the Research Fund for the Doctoral Program of Higher Education (20060055010)
文摘This article presents a novel method to prove that: let E be an AM-space and if dim E ≥ 3, then there does not exist any odd subtractive.isometric mapping from the unit sphere S(E) into S[L(Ω, μ)]. In particular, there does not exist any real linear isometry from E into L(Ω, μ).
文摘针对化工连续生产过程的时序性及非线性等特征,提出一种新的基于数据驱动的化工过程故障诊断方法:ISOMAP-LDA。首先实行流形学习算法ISOMAP,在保持量测数据几何结构特性下完成非线性降维,然后基于提取的嵌入变量张成的低维空间,选用线性判别分析(LDA)构造故障模式类的判别函数,负责各采样个体故障类型的判定。将该方法用于仿真化工Tennessee East man过程的故障诊断,结果表明,ISOMAP-LDA方法不仅拥有较高的故障诊断能力,而且取得采样在低维空间的可视化表示。