基于降维分析的大坝安全特征提取方法及其应用研究

为有效提取大坝安全的低维特征,将特征正交分解、等概率粗粒化、主成分/核主成分分析等降维方法与大坝安全特征提取方法相结合,提出了基于POD的分位数回归分析法、基于实测资料复杂网络分析法、基于PCA/KPCA的预警指标提取法以及基于POD的多场耦合强度折减法等4种新方法,研究表明,上述4种方法能提取高维实测数据或无穷维大坝安全的低维特征,使得大坝安全性态分析和判断可以在低维空间进行,典型工程实例检验了新方法的有效性。

甘肃省武威酿酒葡萄种植区土壤重金属分布特征及来源分析

该试验研究了武威3个酿酒葡萄种植区土壤中8种重金属元素,运用单因子污染指数及内梅罗综合污染指数对酿酒葡萄种植土壤中重金属污染进行评价,并借助主成分分析/绝对主成分分数(principal component analysis/absolute principal component score,PCA/APCS)受体模型对重金属来源进行了解析。结果表明,武威3个酿酒葡萄种植区土壤8种重金属元素的含量均低于国家标准,但以甘肃省土壤背景值为评价依据时,结果显示3个区域均存在轻度污染或少量样点重度污染。PCA/APCS受体模型显示A区主要可分为自然源(包含少量大气沉降源和农业活动源)及农业活动源;B区分为自然源、农业活动源和工业源;C区分为自然源与农业活动结合源以及交通源。农业活动源是各种植区主要重金属污染源,Cd为其特征元素,且各区Cd的不同来源贡献率空间差异较大。此外,本试验所测8种重金属元素含量虽均低于国家标准,但易受人为活动影响且来源复杂,应加强控制,合理耕作,保障土壤环境质量和酿酒葡萄品质,进而保证所生产葡萄酒的品质。

基于改进的PCA-RBFNN过程变量软测量建模及应用

针对过程控制系统中关键变量的软测量建模及应用问题,结合主成分分析法(PCA)和径向基(RBF)神经网络法(RBFNN),提出了改进的PCA-RBFNN软测量建模方法。首先利用PCA分析变量筛选法从过程变量集合中找到对系统过程特性具备最佳解释能力的过程变量子集;然后将该过程变量子集作为输入、被估计变量作为输出构建PCA-RBFNN模型,并使用K-means聚类和最小均方误差法初始化RBF神经网络的数据中心、扩展系数和连接权值;最后采用梯度下降法训练、校正所建模型。以某纺织原料生产过程为实例,对所建模型进行了验证和输出性能对比分析。结果表明,该模型可以实现过程变量在线预测,比原模型具有更好的泛化能力、预测能力和输出精度,能够提高过程控制系统的稳定性和可靠性。

Identification of Heterogeneity of Social and Economic Environment of Land Uses in China

The robust principal component analysis （RPCA） is a technique of multivariate statistics to assess the social and economic environment quality. This paper aims to explore a RPCA algorithm to analyze the spatial heterogeneity of social and economic environment of land uses （SEELU）. RPCA supplies one of the most efficient methods to derive the most important components or factors affecting the regional difference of the social and economic environment. According to the spatial distributions of the levels of SEELU,the total land resources of China were divided into eight zones numbered by Ⅰ to Ⅷ which spatially referred to the eight levels of SEELU.

Construction of Anti-breaking Models of the Main Veins of Flue-cured Tobacco Leaves and Principal Component Analysis

[Objective] This study aimed to explore the related mechanisms of the breaking of flue-cured tobacco leaves. [Method] Anti-breaking models of the main veins of flue-cured tobacco leaves were constructed for principal component analysis on the anti-breaking index, leaf traits and cellulose contents. [Result] The results showed that the growth traits had certain relevance with the cellulose contents while the leaf weight assumed a significant negative correlation with the anti-breaking index, indicating that the heavier the leaf weight was, the weaker the anti-breaking capacity of flue-cured tobacco would be; the cross-sectional area of main veins and the cellulose contents had shown a positive correlation with the anti-breaking index, indicating that the thicker the main vein of flue-cured tobacco was, the higher the cellulose contents would be, and the stronger the anti-breaking capacity of flue-cured tobacco leaves would be. [Conclusion] This study provided theoretical basis and reference to improve tobacco production and enhance the quality of flue-cured tobacco.

北京市春夏挥发性有机物的污染特征及源解析

2017年4月—7月期间对北京北五环附近大气中的挥发性有机物(VOCs)进行观测研究,并利用主成分分析/绝对主因子得分受体模型对VOCs进行源解析,同时结合气象数据推判污染源分布.结果表明,观测期间VOCs主要有烷烃、烯烃、芳香烃、卤代烃和含氧烃等.源解析的结果表明,VOCs来源于燃料挥发、汽车尾气排放、溶剂挥发、单一的溶剂使用、工业排放和干洗六种污染源.其中汽车尾气排放、燃料挥发和溶剂挥发共占所有污染源的73%,是主要的污染源.观测期间污染物的高值区主要分布于观测点南部,而无具体污染源的污染物则可能受气象等因素的影响而更广泛地分布在观测点周围.

海相沉积粉质中间土的物理状态

长江河口地区通启高速公路工程地质研究发现,该区域广泛分布海相粉质中间土,其液性指数与一般海相软粘土相似,具有较高的原位强度、较低的塑性和显著的结构性,而其粒度分布、强度特性、压缩性和塑性图与海相软粘土明显不同.采用主成分与主分量分析,对海相粉质中间土的多指标反映其土性信息的能力进行了分析.分析结果表明,采用液性指数划分其稠度的传统方法不尽合理,而采用F孔隙函数确定其物理状态更加合理.根据物理状态指标与工程力学特性的相关性,建立孔隙函数划分中间土物理状态新方法.通启高速公路工程实践表明,海相粉质中间土具有良好的力学特性,不应隶属于软粘土地基.

Kernel principal component analysis network for image classification

In order to classify nonlinear features with a linear classifier and improve the classification accuracy, a deep learning network named kernel principal component analysis network（ KPCANet） is proposed. First, the data is mapped into a higher-dimensional space with kernel principal component analysis to make the data linearly separable. Then a two-layer KPCANet is built to obtain the principal components of the image. Finally, the principal components are classified with a linear classifier. Experimental results showthat the proposed KPCANet is effective in face recognition, object recognition and handwritten digit recognition. It also outperforms principal component analysis network（ PCANet） generally. Besides, KPCANet is invariant to illumination and stable to occlusion and slight deformation.

Adaptive partitioning PCA model for improving fault detection and isolation

In chemical process, a large number of measured and manipulated variables are highly correlated. Principal component analysis(PCA) is widely applied as a dimension reduction technique for capturing strong correlation underlying in the process measurements. However, it is difficult for PCA based fault detection results to be interpreted physically and to provide support for isolation. Some approaches incorporating process knowledge are developed, but the information is always shortage and deficient in practice. Therefore, this work proposes an adaptive partitioning PCA algorithm entirely based on operation data. The process feature space is partitioned into several sub-feature spaces. Constructed sub-block models can not only reflect the local behavior of process change, namely to grasp the intrinsic local information underlying the process changes, but also improve the fault detection and isolation through the combination of local fault detection results and reduction of smearing effect.The method is demonstrated in TE process, and the results show that the new method is much better in fault detection and isolation compared to conventional PCA method.

基于受体模型和地统计学的海南岛农用地土壤重金属分布特征及源解析

基于海南岛农用地表层土壤中8种重金属含量测定结果,使用多元统计及污染指数法评价重金属污染风险,结合主成分分析/绝对主成分得分受体模型及地统计学方法对重金属污染来源及其分布特征作解析。结果表明:(1)海南岛农用地土壤重金属总体处于较低浓度水平,但存在局部区域富集情况。(2)污染来源主要有自然源、工业交通源及农业源。自然源对Ni、Cr、Cu的贡献率分别为79.1%、60.5%、45.5%,高值集中在海南岛东北部玄武岩覆盖区;工业交通源对Hg、Pb、Zn的贡献率分别为73.6%、73.5%和59.4%,高值散落于环岛高速及港口区域、主要城市、交通枢纽所在及矿产开采区;农业源对Cd、As的贡献率分别为62.9%、54.9%,高值分布于海南岛沿海平原农业密集区。

Contemporary Analytic Philosophy and Bayesian Subjectivism： Why Both Are Incoherent

My purpose in this paper is to argue for two separate, but related theses. The first is that contemporary analytic philosophy is incoherent. This is so, I argue, because its methods contain as an essential constituent a non-classical conception of intuition that cannot be rendered consistent with a key tenet of analytic philosophy unless we allow a Bayesian-subjectivist epistemology. I argue for this within a discussion of two theories of intuition： a classical account as proposed by Descartes and a modem reliabilist account as proposed by Komblith, maintaining that reliabilist accounts require a commitment to Bayesian subjectivism about probability. However, and this is the second thesis, Bayesian subjecfivism is itself logically incoherent given three simple assumptions： （1） some empirical propositions are known, （2） any proposition that is known is assigned a degree of subjective credence of 1, and （3） every empirical proposition is evidentially relevant to at least one other proposition. I establish this using a formal reductio proof. I argue for the t-u-st thesis in section 1 and for the second in section 2. The final section contains a summary and conclusion.

南京北郊大气VOCs变化特征及来源解析

利用2011-03-01~2012-02-29南京北郊大气VOCs观测资料,对大气VOCs浓度变化特征和特征物比值差异展开研究,并应用PCA/APCS受体模型对不同季节VOCs来源进行了解析.结果表明,南京大气总VOCs体积混合比为43.52×10-9,其中烷烃占45.1%、烯烃占25.3%、炔烃占7.3%和芳香烃占22.3%.总VOCs体积混合比呈现夏季高,冬季低的季节变化.VOCs组分中烷烃在冬季最高,烯烃夏季最高,芳香烃春季最高,炔烃冬季最高.特征物比值(VOCs/乙炔)和T/B比值反映出观测点受周边工业区影响较大.VOCs源解析表明,主要来源来自工厂生产、机动车排放、燃料燃烧、生产活动挥发、溶剂使用和自然源.虽然有季节变化,但与工业生产活动相关的来源占大气VOCs 45%~63%,其次为机动车来源占34%~50%.