控制图等多种质量控制方法在大豆粉检测能力验证中的应用

目的对FAPAS 07322大豆粉中铝、总砷、镉、铅和总汞能力验证结果的准确性和可靠性进行验证。方法分别采用方法比对、仪器比对、加标回收试验及绘制控制图等方式进行质量控制。结果方法相对标准偏差(relative standard deviation,RSD)在0.9%～2.4%,仪器比对RSD在0.9%～2.3%,加标回收率在98.9%～107.4%,测定结果均在警告限范围,控制图可靠。能力验证|Z|在0.2～0.4。结论控制图等多种质控方法相结合在能力验证中的应用,可以为同行检验检测机构提供技术参考及方法学依据。     

基于改进雷达图的配电系统电能质量评估方法

本文针对配电系统电能质量综合评估存在的评估指标权重分配过于主观或客观的问题,提出一种基于改进雷达图法的配电系统电能质量评估方法。该方法利用最小偏差组合权重法计算各评估指标的最优组合权重。并在传统雷达图的基础上对各指标轴间夹角值的确定和特征向量的选取两方面做了进一步的改进。首先采用扇形区域取代原始的三角形区域,并利用最小偏差组合权重法计算各指标轴间夹角值。其次选择新的特征向量构造评估函数对配电系统电能质量进行综合评估。最后通过对某钢铁厂棒材变电站10kV配电系统电能质量进行综合评估,验证了该方法的可行性和合理性。较传统雷达图而言,改进雷达图法解决了传统雷达图法中各指标间信息共用以及评估结果不唯一的问题,使评估结果更加准确合理。     

An EWMA signed ranks control chart with reliable run length performances

During the design phase of a control chart, the determination of its exact run length properties plays a vital role for its optimal operation. Markov chain or integral equation methods have been extensively applied in the design phase of conventional control charts. However, for distribution-free schemes, due to the discrete nature of the statistics being used (such as the sign or the Wilcoxon signed rank statistics, for instance), it is impossible to accurately compute their run length properties. In this work, a modified distribution-free phase II exponentially weighted moving average (EWMA)-type chart based on the Wilcoxon signed rank statistic is considered and its exact run length properties are discussed. A continuous transformation of the Wilcoxon signed rank statistic, combined with the classical Markov chain method, is used for the determination of the average run length in the in- and out-of control cases. Moreover, its exact performance is derived without any knowledge of the distribution of sample observations. Finally, an illustrative example is provided showing the practical implementation of our proposed chart.     

A side-sensitive synthetic coefficient of variation chart

Control charts for monitoring the coefficient of variation (γ) are useful for processes with an inconsistent mean (μ) and a standard deviation (σ) which changes with μ, by monitoring the consistency in the ratio σ over μ. The synthetic-γ chart is one of the charts proposed to monitor γ, and its attractiveness lie in waiting until a second point to fall outside the control limits before a decision is made. However, existing synthetic-γ charts do not differentiate between the points falling outside the upper control limit (UCL) and lower control limit (LCL). Hence, this paper proposes a side-sensitive synthetic-γ chart, where successive nonconforming samples must either fall above the UCL or below the LCL. Formulae to compute the average run length (ARL), the standard deviation of the run length (SDRL) and expected average run length (EARL) are derived using the Markov chain approach, and the algorithms to obtain the optimal charting parameters are proposed. Subsequently, the optimal charting parameters, ARL, SDRL and EARL values for various numerical examples are shown. Comparisons show that the side-sensitive synthetic-γ chart consistently outperforms the existing synthetic-γ chart, especially for small shifts. The proposed chart also consistently outperforms the Shewhart-γ chart, while showing comparable or better performance than the Exponentially Weighted Moving Average (EWMA) chart for most shift sizes, except for very small shifts. Finally, this paper shows the implementation of the proposed chart on an industrial example.     

Two successive occasions resubmitted sampling scheme-based control chart

In this paper, a resubmitted sampling-based successive sampling over two successive occasions control chart is proposed to monitor the underlying characteristic of interest. Auxiliary information of the first occasion is utilized to monitor the relative change in the study variable over the second occasion successively despite high degree of correlation. The structural and operational design is presented along with the comparative performance evaluation. The average run length is used as a performance evaluation measure and proved the argument in favor of the presented concept in comparison with the other auxiliary data control charts. The implementation is explained through two real examples.     

一种基于雷达图表示的数值型数据的CNN分类方法

卷积神经网络（convolutional neural networks， CNN）是一种广泛用于分析视觉图像的分类方法.由于数值数据存在着非线性、耦合性等复杂的空间关系，因此基于CNN的数值型数据的研究较少.本文的目的是找到一种可行的方法，将CNN的应用领域扩展到数值数据.于是提出了一种基于雷达图表示的数值型数据的CNN分类方法（Radar-CNN）.该算法首先将数值数据表示成雷达图形式，然后将其输入CNN中构建分类模型.为了进一步研究特征尺度和序列对性能的影响，提出了两种改进算法Rank Radar-CNN和SFS Radar-CNN.为了验证所提算法的有效性，引入TE化工过程数据集进行实验测试并比较，实验结果表明Radar-CNN及其改进算法具有优异的性能.     

New sampling strategies to reduce the effect of autocorrelation on the synthetic T2 chart to monitor bivariate process

In this paper, synthetic T² chart is developed to monitor bivariate process with correlated variables and autocorrelated observations. The proposed chart is a combination of the Hotelling's T² chart and the conforming run length chart. The operation and design of the chart are described when observations are autocorrelated and cross correlated. The first‐order vector autoregressive process VAR (1) is used to model the bivariate data from an autocorrelated process of interest. Using an average run length as performance measure criterion in the VAR (1) model, it is observed that autocorrelation seriously impact the performance of the synthetic T² chart. To reduce the effect of autocorrelation on the performance of the synthetic T² chart, the skip and mixed sampling strategies are implemented to form rational subgroups in the construction of synthetic T² chart. The average run length performance of the synthetic T² chart implementing these strategies is compared with that of the standard strategy of formation of rational subgroups. It is observed that implementing skip and mixed sampling strategies within rational subgroup improves the performance of the synthetic T² chart.     

An artificial neural network approach to inhomogeneous soil slope stability predictions based on limit analysis methods

The assessment of soil slope stability is an important task in geotechnical designs. This study uses finite element upper bound (UB) and lower bound (LB) limit analysis (LA) methods to investigate inhomogeneous soil slope stability on the basis of the conventional Mohr–Coulomb parameters. The obtained stability numbers are presented in inhomogeneous soil slope stability charts. In order to minimize manual reading errors when using the chart solutions, an artificial neural network (ANN) is employed to develop a stability assessment tool for the slopes investigated in this paper. The slope stability analysis using the ANN-based tool is convenient, and the predictions it provides are highly accurate.     

A one-class peeling method for multivariate outlier detection with applications in phase I SPC

In phase I of statistical process control (SPC), control charts are often used as outlier detection methods to assess process stability. Many of these methods require estimation of the covariance matrix, are computationally infeasible, or have not been studied when the dimension of the data, $p$, is large. We propose the one-class peeling (OCP) method, a flexible framework that combines statistical and machine learning methods to detect multiple outliers in multivariate data. The OCP method can be applied to phase I of SPC, does not require covariance estimation, and is well suited to high-dimensional data sets with a high percentage of outliers. Our empirical evaluation suggests that the OCP method performs well in high dimensions and is computationally more efficient and robust than existing methodologies. We motivate and illustrate the use of the OCP method in a phase I SPC application on a $N=354$, $p=1917$ dimensional data set containing Wikipedia search results for National Football League (NFL) players, teams, coaches, and managers. The example data set and R functions, OCP.R and OCPLimit.R, to compute the respective OCP distances and thresholds are available in the supplementary materials.