FRACTAL ANALYSIS OF TECHNICAL PARAMETERS FOR FORECASTING TASKS
Abstract
The main objective of the monitoring system of a technical object is to monitor the values of
its parameters at the current time and to predict possible changes in parameters for a given time
period. The assessment of parameter changes for the period of time for which the forecast is calculated
is based on the processing of previously obtained data. The horizon and forecasting accuracy
depend on the quality of the time series of data used for the calculations. The nature of the data
series also determines the choice of a specific forecasting model, the need and the choice of themethod of their preliminary processing. The more unstable the trend of the parameter’s time series,
the less possibilities there are for estimating the parameter values outside the current time
frame. Trend resistance of a series is determined using the normalized range method. The result of
using the method is to assess the nature of the time series as a whole. However, most methods of
forecasting time series use a limited part of the series directly adjacent to the time point of the
beginning of the forecasting process. And the nature of this section affects the accuracy of the
forecast. If this part of the series can be defined as persistent, then a forecast is possible. To analyze
the parameter values belonging to a small part of the series, the normalized span method is
not suitable, since it requires a large amount of data to be sampled. In this case, you can use the
fractality index, which allows local analysis. This paper presents a phased fractal analysis of the
time series data of a technical object parameter. At the first stage, a preliminary analysis is carried
out using the normalized scale method. The nature of the series as a whole is determined.
If the series is persistent, then the next step is the local fractal analysis of a limited area of a series
of parameter values. The possibility of using this section of the series for the implementation of the
forecasting problem is estimated. A local fractal analysis can be performed on a site limited by a
fixed time window that moves at each forecasting step following this process. A similar forecasting
support scheme can improve the accuracy and reliability of the forecast.
References
of Parameters // 2016 International Siberian Conference on Control and Communications,
SIBCON 2016 - Proceedings. – 2016. – P. 749-752.
2. Detlev W. Gross Partial Discharge Measurement and Monitoring on Rotating Machines //
IEEE Int. Sym. On Elect. Insul, Boston MAUSA, April 7-10, 2002. – P. 33-41.
3. Клевцова А.Б., Клевцов Г.С. Модели параметрической экспресс-оценки состояния тех-
нического объекта // Известия ЮФУ. Технические науки. – 2008. – № 11 (88). – С. 15-19.
4. Клевцов С.И. Использование моделей временных рядов для краткосрочного прогнози-
рования в микроконтроллере изменений параметров объекта // Известия ЮФУ. Техни-
ческие науки. – 2013. – № 11 (148). – С. 194-201.
5. Lihua Sun, Yingjun Guo, Haichao Ran. A New Method of Early Real-Time Fault Diagnosis
for Technical Process // Electrical and Control Engineering (ICECE), 2010 International Conference.
– Wuhan, China, 2010. – P. 4912-4915.
6. Darkhovsky B., Brodsky B. Asymptotically Optimal Methods of Early Change-point Detection
// Sequential Analysis. – 2013. – No. 32. – P. 158-181.
7. Федер Е. Фракталы: пер. с англ. – М.: Мир, 1991. – 254 с.
8. Антипов О.И., Неганов В.А. Применение метода нормированного размаха Хёрста к ана-
лизу стохастических временных рядов в импульсных стабилизаторах напряжения // Фи-
зика волновых процессов и радиотехнические системы. – 2009. – Т. 12, № 3. – С. 78-85.
9. Matuszewski J. Application of clustering methods for recognition of technical objects // Modern
Problems of Radio Engineering, Telecommunications and Computer Science (TCSET),
2010 International Conference. – 2010. – P. 39-40.
10. Roel F. Ceballos, Fe F. Largo On. The Estimation of the Hurst Exponent Using Adjusted Rescaled
Range Analysis, Detrended Fluctuation Analysis and Variance Time Plot: A Case of
Exponential Distribution // Imperial Journal of Interdisciplinary Research (IJIR). – 2017.
– Vol. 3, Issue 8. – P. 424-434.
11. F. Cervantes-de la Torre, Jesús Isidro González-Trejo, Cesar Augusto Real-Ramirez, Luis F.
Hoyos-Reyes. Fractal dimension algorithms and their application to time series associated with
natural phenomena // Journal of Physics: Conference Series. – 2013. – No. 475. – P. 1-10.
12. Klevtsov S. Using the Method of Normalized Amplitude for Assessing the Quality of the Calibration
Tests of the Pressure Sensor // 2019 Ural Symposium on Bio-medical Engineering, Radioelectronics
and Information Technology (USBEREIT). – Yekaterinburg, Russia, 2019. – P. 197-199.
13. Box George E.P., Box, Gwilym M. Jenkins, Gregory C. Reinsel. Time series analysis: forecasting
and control. – 4th ed. – A John Wiley & Sons, Inc., Publication, 2015. – 712 p.
14. Кириченко Л., Чалая Л. Комплексный подход к исследованию фрактальных временных
рядов // International Journal "Information Technologies & Knowledge". – 2014. – Vol. 8,
No. 1. – P. 22-28.
15. Калуш Ю.А., Логинов В.М. Показатель Хёрста и его скрытые свойства // Сиб. журн. ин-
дустр. матем. – 2002. – Т. 5, № 4. – С. 29-37.
16. Биченова Н. Вычисление показателя Херста для динамики стоимости компании // Automated
Control Systems. Transactions. Georgian Technical University. – 2015. – No. 1 (19).
17. Кузенков Н.П., Логинов В.М. Использование метода нормированного размаха при анали-
зе речевых патологий неврологического генеза // Компьютерные исследования и моде-
лирование. – 2014. – Т. 6, № 5. – С. 775-791.
18. Бельков Д.В., Едемская Е.Н., Незамова Л.В. Статистический анализ сетевого трафика //
Наукові праці ДонНТУ. Серія "Інформатика, кібернетика та обчислювальна техніка".
– 2011. – Вип. 13 (185). – С. 66-75.
19. James B. Bassingthwaighte, Gary M. Raymond. Evaluation of the Dispersional Analysis
Method for Fractal Time Series // Ann Biomed Eng. – 1995. – Vol. 23 (4). – P. 491-505.
20. Дубовиков М.М., Крянев А.В., Старченко Н.В. Размерность минимального покрытия и
локальный анализ фрактальных временных рядов // Вестник РУДН. – 2004. – Т. 3, № 1.
– С. 81-95.
21. Старченко Н.В. Индекс фрактальности и локальный анализ хаотических временных рядов с
помощью индекса фрактальности: автореф. дисс. … канд. физ.-мат. наук. – М., 2005.
22. Васильев В.В. Вычисление индекса фрактальности временного ряда // Вестник ТГУ.
– 2011. – Т. 16. – Вып. 4. – С.1047-1049.
23. Белолипцев И.И., Фархиева С.А. Предсказание финансовых временных рядов на основе
индекса фрактальности // Мир науки. – 2014. – № 3. – С. 1-12.
24. Владимирова Д.Б. Индекс фрактальности в исследованиях детерминированности дис-
кретных временных рядов // Science and Business: Development Ways. – 2015. – No. 8
(50). – P. 86-91.
25. Dubovikov M.M., Starchenko N.V. and Dubovikov M.S. Dimension of the minimal cover and
fractal analysis of time series // Physica A: Statistical Mechanics and its Applications. – 2004.
– Vol. 339, Issues 3-4. – P. 591-608.
