Treatment of missing values in process data analysis

Process data suffer from many different types of imperfections. For example, bad data due to sensor problems, multi‐rate sampling, outliers, compressed data etc. Since most modelling and data analysis methods are developed to analyze regularly sampled and well conditioned data sets there is a need for pre‐treatment of data. Traditionally data conditioning or pre‐treatment has been done without taking into account the end use of the data, for example, univariate methods have been used to interpolate bad data even when the intended end use of data is for multivariate analysis. In this paper we consider the pre‐treatment and data analysis as a collective problem and propose data conditioning methods in a multivariate framework. We first review classical process data analysis methods and acclaimed missing data handling techniques used in statistical surveys and biostatistics. The applications of these acclaimed missing data techniques are demonstrated in three different instances: (i) principal components analysis (PCA) is extended in data augmentation (DA) framework for dealing with missing values, (ii) iterative missing data technique is used to synchronize uneven length batch process data, and (iii) PCA based iterative missing data technique is used to restore the correlation structure of compressed data.     

基于TGNPE算法的间歇过程故障诊断

间歇过程数据是由批次、变量和时间构成的三维数据,数据内包含了丰富的对过程监控有用的全局和局部结构信息,如何充分提取间歇过程的特征信息是故障诊断的关键。传统方法处理三维数据都是将其展开成二维数据,展开过程必然会导致数据内在结构破坏,并且通常只考虑了数据的全局信息或者只考虑了数据的局部信息,这就不能充分提取过程的有用信息导致诊断效果欠佳。针对以上问题,提出了张量全局-局部邻域保持嵌入(TGNPE)算法,首先用张量分解的方法直接对三维数据进行建模,而不对数据进行展开,这就有效地保存了数据的内部结构,再用邻域保持嵌入算法充分提取数据局部结构信息的同时兼顾数据的全局信息,这就实现了对数据特征信息更加充分地提取,用TGNPE算法检测到故障后用贡献图法诊断出故障变量。通过青霉素发酵过程验证了本文提出的算法对间歇过程数据信息提取更加充分,更利于故障诊断。     

EDF调度方法在交换式工业以太网中的实现

针对工业以太网的数据特点 ,通过在以太网数据帧中引入IEEE80 2 .1Q优先级字段来区分周期性实时数据和非实时数据 ,并对前者优先进行传送。同时还在周期性数据帧中引入截止期限字段来区分其不同的紧急程度 ,并对其采用非中断的EDF调度方法来改善周期性数据帧的端到端延迟。最后通过仿真试验验证了该方法的可行性。     

The role of proxy information in missing data analysis

This article investigates the role of proxy data in dealing with the common problem of missing data in clinical trials using repeated measures designs. In an effort to avoid the missing data situation, some proxy information can be gathered. The question is how to treat proxy information, that is, is it always better to utilize proxy information when there are missing data? A model for repeated measures data with missing values is considered and a strategy for utilizing proxy information is developed. Then, simulations are used to compare the power of a test using proxy to simply utilizing all available data. It is concluded that using proxy information can be a useful alternative when such information is available. The implications for various clinical designs are also considered and a data collection strategy for efficiently estimating parameters is suggested.     

流程工业数据采集中的流量控制

数据采集系统对数据库服务器需要进行非常频繁地读写,因此要考虑进行流量控制,尽可能地减少通信流量。提出一种group机制,按照已定义的数据点顺序直接发送其数据值序列,而不再重复传输数据点编号信息。通过优化传统数据传输方式中位号点信息（如节点编号）的传送来大幅减少数据流量,从而提高数据传输速度,更突出实时性。研究中使用DTD对数据请求和应答进行规范化定义,并使用xm l文档进行数据传递示例说明。group方式最适用于在数据点数目比较多,所属数据点变化不是很频繁的情况下,它能减少至少一半的网络流量。     

基于两步矩阵投影的数据分类算法

提出了一种数据分类的两步矩阵投影算法.指出Crowe提出的矩阵投影算法在数据分类中存在由于投影矩阵不惟一,导致已测可校正数据分类不彻底的缺点.采用已测数据预分类的方法,对其进行了修正.在此基础上,将矩阵投影算法引入到了未测数据分类中,提出了基于矩阵投影算法的未测数据分类算法.新算法只需求解两个投影矩阵就可以实现所有数据分类.从而避免了常规方法在未测数据分类时,求解未测数据关联矩阵绝对线性无关列的计算,提高了计算效率.数学推导和算例验证了新算法的有效性.     

基于动态单类随机森林的非线性过程监控方法

针对高维化工过程中存在的非线性和动态特性，提出了一种基于动态单类随机森林（dynamic one-class random forest，DOCRF（的过程监控方法。对正常运行状态下的过程数据进行稀疏性分析，根据其反分布产生离群点数据。利用典型变量分析对正常数据进行相关性分析，分别将正常数据和离群点数据投影到典型变量空间，利用典型变量空间数据训练单类随机森林。基于单类随机森林模型根据待检测样本与正常数据的相似度构造监控统计量进行故障检测。在Tennessee Eastman过程的仿真结果表明，所提DOCRF方法总体优于单类支持向量机方法。     

氯离子对循环水中COD的测试的影响及掩蔽

COD的准确测试在日常监测中对判断循环水系统的有机物泄漏、保证循环水微生物控制和正常供水运行都具有现实的指导作用。通过日常的监测数据统计分析,得出了COD测试对于循环水日常微生物控制的重要意义,油含量和COD二者所表征的有机物浓度具有较大差异,同时指出,氯根在50~500 mg·L-1浓度下的循环水中仍旧与COD测试数据具有良好线性相关性,需要进行必要的修正和排除干扰。硝酸银掩蔽试验数据表明,循环水COD测试时需要根据循环水水样的氯根浓度加入不同量的掩蔽剂,才能得到真实有效数据。     

Evaluation of software for multiple imputation of semi-continuous data

It is now widely accepted that multiple imputation (MI) methods properly handle the uncertainty of missing data over single imputation methods. Several standard statistical software packages, such as SAS, R and STATA, have standard procedures or user-written programs to perform MI. The performance of these packages is generally acceptable for most types of data. However, it is unclear whether these applications are appropriate for imputing data with a large proportion of zero values resulting in a semi-continuous distribution. In addition, it is not clear whether the use of these applications is suitable when the distribution of the data needs to be preserved for subsequent analysis. This article reports the findings of a simulation study carried out to evaluate the performance of the MI procedures for handling semi-continuous data within these statistical packages. Complete resource use data on 1060 participants from a large randomized clinical trial were used as the simulation population from which 500 bootstrap samples were obtained and missing data imposed. The findings of this study showed differences in the performance of the MI programs when imputing semi-continuous data. Caution should be exercised when deciding which program should perform MI on this type of data.     

Estimating Tensile Creep Data from Flexure Data

Safety factors are derived for converting bend data to tensile data for creep. It is assumed that the ceramic being stressed is subject to preferential damage in tension. The analysis shows that the safety factors can be relaxed considerably when compression data are available to augment the bend data.     

A new method of processing the time-concentration data of reaction kinetics

Experimental data of reaction kinetics are usually in the form of concentration versus time. For kinetics investigation it is more convenient to have the data in the form of reaction rate versus concentration. Converting time-concentration data into concentration-reaction rate data is an ill-posed problem in the sense that if inappropriate methods are used the noise in the original data will be amplified leading to unreliable results. This paper describes a conversion procedure, independent of reaction rate model or mechanism, that manages to keep noise amplification under control. The performance of this procedure is demonstrated by applying it to several sets of published kinetic data. Since these data are accompanied by their rate equations, the computed rates are used to obtain the unknown parameters in these equations. Comparison of these parameters with published figures and the ease with which they are obtained highlights the advantages of the new procedure.     

Master curve approach to polymer crystallization kinetics

Nonisothermal crystallization kinetic data obtained from differential scanning calorimetry (DSC) for a poly(ethylene terephthalate) are corrected for the effects of temperature lag between the DSC sample and furnace using the method of Eder and Janeschitz-Kriegl which is based on experimental data alone without resort to any kinetic model. A method is presented for shifting the corrected nonisothermal crystallization kinetic data with respect to an arbitrarily chosen reference temperature to obtain a master curve. The method is based on experimental data alone without reference to any specific form of kinetic model. When the isothermal crystallization kinetic data for the same material are shifted with respect to the same reference temperature, a master curve is also obtained which overlaps to a large extent the corresponding master curve from nonisothermal data. It follows that nonisothermal DSC measurements provide the same crystallization kinetic information as isothermal DSC Measurements, only over a wider range of temperatures. The shift factors obtained from experimental data alone are compared in turn with the corresponding values calculated from the Avrami equation, the Hoffman-Lauritzen expression, and the Nakamura equation as a means of evaluating these models individually. It is concluded that the Avrami equation is very good at describing isothermal crystallization kinetics, the Hoffman-Lauritzen extrapolation of the limited isothermal data to a wide range of temperatures is quite good, and the Nakamura equation yields reliable crystallization kinetic information over a narrower range of temperatures than nonisothermal data alone without using any specific model.     

CIPS中的数据仓库技术

数据仓库技术是综合利用关系复杂的海量数据的有效手段。本提出数据仓库技术应用到ＣＩＰＳ中由于一过大、数据过于复杂、历史数据过多而造成的事实上无法有效使用的海量数据变得更有价值,在过程生产方面发挥列在的作用。首先概括性地描述了数据仓库的基本要领与实现方法,在此基础上阐述了在ＣＩＰＳ中建立数据仓库的意义、数据仓库在ＣＩＰＳ中的重要作用和地位、ＣＩＰＳ中实现数据仓库的方法以及建立ＣＩＰＳ数据仓库的步骤。     

苯、甲苯和对二甲苯二元和三元系的汽液平衡

本文在改进的Dvorak-Boublik平衡釜中测定了101.33kPa下的苯、甲苯和对二甲苯三组二元系和一组三元系的汽液平衡(VLE)数据。对二元的实验数据作了热力学一致性检验和Wilson方程关联,又由回归二元数据得到的Wilson方程参数推算了三元VLE数据,并将其与测定的数据作了比较,两者符合得较好。     

用粗糙集理论确定矿井通风经验公式的方法

运用数据挖掘中的粗糙集理论确定在定性和定量因素均存在的条件下矿井通风经验公式的原理 ,给出了定性和定量数据融合、冗余数据和无用因素的清除、数据间的相对重要性的确定以便如何采集数据等一系列方法     

水力模型在上海城市原水系统的实践与应用

文中拟通过应用Infoworks WS Pro和Infoworks ICM这两种软件模拟上海城市原水输水系统,比较两种软件在模拟大型城市原水系统的优劣。主要针对典型案例进行模拟,通过采用两种软件的不同糙率模型,对不同坡度及管底标高的微观模型进行测试。结果发现,WS Pro和ICM模型的计算原理及建模数据要求不同,WS Pro对管道数据相对于ICM不敏感,ICM的CW糙率模型在基于真实的GIS模型之上能获得最优解。主要结论为:对于无精细管道数据的原水系统,推荐采用WS Pro模型的CW糙率模型;对于有详细管网数据的原水系统,推荐采用ICM的N糙率模型。     

基于重构的半监督ELM及其在故障诊断中的应用

工业过程中获取带标签的故障数据困难,而无标签故障数据却大量存在,如何有效地利用数据信息进行故障诊断是故障诊断领域的重要内容。为更充分地挖掘和利用数据信息,提出一种新的半监督学习方法：基于重构的半监督极限学习机（RSELM）。相比于传统的半监督极限学习机（ELM）方法,RSELM采用自动编码ELM（ELM-AE）获得的输出权重替代随机的隐含层输入权重,能更有效地提取数据特征;考虑到数据均可由其近邻数据来线性重构,故可构建近邻数自适应选择的重构图,并同时利用数据的标签信息优化连接权重,以更优地反映数据结构信息;通过建立新的含局部保持的目标函数,可有效地训练分类器。标准数据集和TE过程上的仿真实验验证了所提算法的有效性。     

一种邮件服务器日志数据分析方法

网络日志及时、准确地记录了网络服务器的运行、服务信息,是网络管理人员应该重点关注的信息源。由于信息量巨大,仅凭人工查阅是无法准确分析系统所提供的信息。建议一种利用统计、数据挖掘技术对邮件日志服务器的分析方法,着重对邮件服务器事件进行时间分布统计、周期模式挖掘和孤立点的分析。     

Theoretical interpretation of gas oil conversion data on an x-sieve catalyst

A four parameter model based on the catalyst aging theory of Wojciechowski⁽⁵⁾ is applied to experimental conversion data in gas oil cracking. The data extends over a range of cat/oil ratios from 0.094 to 3.0, and run times from 1.25 to 40 minutes. Within this range of conditions the model is applied to data with and without thermal conversion and excellent correlation between experimental results and calculated values is obtained. The technique for fitting data to the model is described in some detail.