Technical gazette, Vol. 26 No. 4, 2019.
Original scientific paper
https://doi.org/10.17559/TV-20190607101917
Identification of Extreme Temperature Fluctuation in Blast Furnace Based on Fractal Time Series Analysis
Shihua Luo
orcid.org/0000-0002-0793-7950
; School of Statistics, Jiangxi University of Finance and Economics, 169 East Shuanggang Revenue, Nanchang, Jiangxi, China
Zian Dai
orcid.org/0000-0002-6759-5761
; School of Statistics, Jiangxi University of Finance and Economics, 169 East Shuanggang Revenue, Nanchang, Jiangxi, China
Fan Guo
; School of Statistics, Jiangxi University of Finance and Economics, 169 East Shuanggang Revenue, Nanchang, Jiangxi, China
Jiusun Zeng
; School of Measurement and Testing Engineering, China Jiliang University, Hangzhou, China
Abstract
In this study, we aim to estimate the density distribution for the return intervals of extreme temperature fluctuation in blast furnace during iron making process. We first identified the fractal feature of the data based on R/S analysis and also calculated the Hurst coefficient. Secondly, based on the fractal feature of the data, we estimated a stretched exponential distribution of the return intervals of extreme temperature fluctuation. Finally, in order to test the result, we applied this method to the data of two blast furnaces, and compared with the traditional kernel density estimation method. The comparison was based on 100,000 times K-S test. The comparison results showed that the fractal time series estimation provides a greater fitness than traditional estimation method since it has no rejection in K-S test. With this method, the density of return intervals of unexpected temperature fluctuation can be estimated. This can be applied as a tool of temperature control and also can be applied as a tool to evaluate the efficiency of the control system.
Keywords
fractal time series; R/S analysis; return intervals; stretched exponential distribution
Hrčak ID:
223309
URI
Publication date:
25.7.2019.
Visits: 1.631 *