文献类型：期刊文献

英文题名：Systematic analysis of chemical dangerous product leakage based on large data technology

作者：Zhang, Qin[1]; Liu, Yutang[2]

第一作者：张秦

通讯作者：Zhang, Qin

机构：[1] School of Mathematics and Information Science, Xinxiang University, Xinxiang, Henan, 453000, China; [2] Department of Basic Subjects, Henan Institute of Technology, Xinxiang, 453002, China

第一机构：新乡学院数学与信息科学学院

年份：2018

卷号：71

起止页码：1027-1032

外文期刊名：Chemical Engineering Transactions

收录：EI(收录号：20190206367747);Scopus(收录号：2-s2.0-85059776141)

语种：英文

外文关键词：Big data - Chemical stability - Signal to noise ratio

摘要：Due to facility upsizing of the chemical system itself, the high process continuity and the complicated interparameter mechanism, the chemical system has its own characteristics in addition to the above 4V (Volume, Variety, Velocity, High Value) features: High dimensionality, strong non-linearity, unevenly distributed sample data, low signal to noise ratio. It is due to these unique features of chemical system that there have some difficulties in the analysis and mining of its big data from the traditional methods. In this paper, Chemical characteristics of system is constructed based on big data platform architecture, and uses the hybrid diagnostic identification algorithm, preidentification of abnormal state of chemical system that may arise, to prevent and avoid the effect of practical application. By analyzing the error based on exception identification method of single point parameter, we can see that the method of abnormal identification of chemical systems based on large data technology has high stability and reliability. ? 2018, AIDIC Servizi S.r.l.