
Main / Iurii Andreevich Kravtsov
Author: "Iurii Andreevich Kravtsov"
Vladimir Nikolaevich Klyachkin, Ulyanovsk State Technical University, Doctor of Engineering; graduated from the Faculty of Mechanics at Ulyanovsk Polytechnic Institute; Professor at the Department of Applied Mathematics and Informatics at Ulyanovsk State Technical University; an author of scientific papers in the field of reliability and statistical methods. [email: v_kl@mail.ru]V. Klyachkin, Iurii Andreevich Kravtsov, Ulyanovsk State Technical University, graduated from the Faculty of Physics and Technology at Ulyanovsk State University, PostGraduate Student at the Department of Applied Mathematics and Informatics at Ulyanovsk State Technical University; an author of scientific papers in the field of statistical process control. [email: ukravtsov@rambler.ru]I. Kravtsov, Dmitrii Anatolievich Zhukov, Ulyanovsk State Technical University, a fifthyear student of a specialty in Applied Mathematics at Ulyanovsk State Technical University; his research interests  statistical diagnostic methods of technical objects. [email: zh.dimka17@mail.ru]D. Zhukov


Evaluation of Object Status Diagnosing Efficiency to Nonrandom Structures Existance on the Hotelling’s Chart
The status of a technical object is often considered emergency if the stability of its operation is broken. multivariate Hotelling’s charts can be used for diagnostics of stability disruption on set of the correlated parameters. The basic criterion of the process stability in this case is the absence of the points above control limit on the chart. However, this criterion does not allow to find out process disruption promptly. For the purpose of increasing sensitivity of the chart to possible failures, it is offered to reveal the presence of nonrandom structures on the chart (the arrangement of points that indicates the process disruption). One more possible way to increase the effectiveness of the control is to use the warning border (the several points that hit in a range between warning and control limits and testify to the process disruption).The effectiveness of these methods is investigated analytically and with the use of statistical tests. The sensitivity substantial increase to possible process disruption is shown by example of the concrete technological process of manufacturing a cover of the sensor of aerodynamic corners. Process stability, hotelling's chart, nonrandom structures, warning border, efficiency of the control.



Sections: Mathematical modeling
Subjects: Mathematical modeling. 
Vladimir Nikolaevich Klyachkin, Ulyanovsk State Technical University, Doctor of Engineering; graduated from the Faculty Mechanics at Ulyanovsk Polytechnic Institute; Professor at the Department of Applied Mathematics and Informatics at Ulyanovsk State Technical University; an author of scientific papers in the field of reliability and statistical methods. [email: v_kl@mail.ru]V. Klyachkin, Tatiana Igorevna Svyatova, Ulyanovsk State Technical University, graduated from the Faculty of Economics and Mathematics at Ulyanovsk State Technical University; PostGraduate Student at the Department of Applied Mathematics and Informatics at Ulyanovsk State Technical University; an author of scientific papers in the field of the statistical process control. [email: tatyana.krasko@rambler.ru]T. Svyatova, Iurii Andreevich Kravtsov, Ulyanovsk State Technical University, graduated from the Faculty of Physics and Technology at Ulyanovsk State University; PostGraduate Student at the Department of Applied Mathematics and Informatics at Ulyanovsk State Technical University; an author of scientific papers in the field of statistical process control. [email: ukravtsov@rambler.ru]I. Kravtsov


Workflow Data Modeling for Analysis of Multidimensional Statistical Control Effectiveness
The multivariate statistical control of the technological process is usually spent with the use of the Hotelling’s chart. If necessary, the generalized dispersion card is used in the process control disruption. Efficiency of the control is evaluated by means of average run length. It is the number of observations from the moment of the process disruption till the moment of the detection of this disruption.For the purpose of increasing the sensitivity of the control to possible disruptions various approaches are used: the analysis of the special kind of structures on the chart, application of the warning border, modification of charts on the basis of algorithms of cumulative sums and exponentially weighted moving averages and others. Thus the average run length sometimes can be determined analytically and more often it has to be estimated with the results of statistical tests. To carry out such tests the sequence of data vectors of technological process is modelled and also various kinds of disruptions are set. Data modeling is offered to implement on the basis of the multivariate normal disruption with the subsequent parameters check. Spasmodic displacement of an average process level and its trend and also spasmodic and gradual increase of the process dispersion are guarded as the basic process disruptions. Multivariate control chart, statistical tests, average run length, sequence of data, process disruption.



Sections: Mathematical modeling
Subjects: Mathematical modeling. 
