scholarly journals APPLICATION OF MARKOV CHAINS TO THE ANALYSIS OF BLAST FURNACE OPERATION EFFICIENCY

2018 ◽  
Vol 61 (8) ◽  
pp. 649-656
Author(s):  
S. K. Sibagatullin ◽  
A. S. Kharchenko ◽  
L. D. Devyatchenko
Keyword(s):  
Markov Chain ◽  
Blast Furnace ◽  
Coke Consumption ◽  
Efficiency Index ◽  
Time Interval ◽  
Short Term ◽  
Efficient Operation ◽  
Term Forecast ◽  
Markov Chain Method ◽  
Discrete States

The article presents the results of modeling in a dynamic format of one of the most important parameters of any research object – the efficiency of its work. As the object of investigation, a blast furnace with a volume of 2014 m3 was chosen. The main parameters of the efficiency of this object are traditionally used daily productivity and specific consumption of coke; these two parameters were generalized in this paper. In this case, various algebraic signs of the influence of these parameters were taken into account in the generalized efficiency index. Taking into account the variation of each of these parameters at 3 levels, the number of levels of the generalized efficiency index was determined as 32 = 9, therefore it was rational to take a 9-point scale with the measuring scale of profitability from the efficient operation of the blast furnace. The two-dimensional array of primary data of the volume N = 177 was transformed into a 9×9 transitional matrix for processing of random transitions of the efficiency index from one state to another by the Markov chain method with discrete states and time. The set of parameters of the random process is calculated: for the long-term forecast – the stationary vector of state probabilities, the average time of recurrence (reversal) for each efficiency state, the evaluation of the blast furnace efficiency in points; for a short-term forecast – the first time of transition from each state to any other state, the step number for a “burst” of probability for each reliable state at the initial moment of time, and the components of the efficiency index are obtained. It was established that the average level of the analyzed efficiency of the blast furnace (daily output 3702 tons and specific coke consumption 470 kg/ton) is achieved mainly due to short-term transitions of low-efficiency states to high-efficiency states and vice versa. The transfer of the system to more efficient and prolonged conditions is possible, and as practice has shown on the same blast furnace after repair works to eliminate the distortion of the furnace profile, the daily productivity has increased to 5048 tons with a specific coke consumption of 445 kg/t, but the structure of the transition matrix and the calculated indicators of the Markov chain have fundamentally changed in the direction of increasing the probabilities of stay and transitions of the system in more efficient states. The use of the Markov chain method with discrete states and time makes it possible to estimate the probable value of the change in the parameters of the operation of a blast furnace in a given time interval with constant levels of parameters characterizing the conditions of its operation.

scholarly journals Using fuzzy time series with and without markov chain: to forecast of edible bird nest exported from Indonesia

2022 ◽  
Vol 335 ◽  
pp. 00016
Author(s):  
Osfar Sjofjan ◽  
Danung Nur Adli
Keyword(s):  
Time Series ◽  
Markov Chain ◽  
Secondary Data ◽  
Fuzzy Time Series ◽  
Percentage Error ◽  
Mean Square ◽  
Forecast Method ◽  
Short Term ◽  
Bird Nest ◽  
Markov Chain Method

Edible bird nest (EBN) were traditional medicine consumed by the Tiongkok. This study compared two-algorithm method. Fuzzy time series and Markov chain as forecast method the number of bird nest exported from Indonesia. The secondary data between 2012 and 2018 were from Bureau Central Statistic (BPS). The scope using in this study were bird nest between 2012 until 2018, with a unit of volume kilograms (Kg). Used secondary export data, collected from BPS of Indonesia. Data were analysed using Fuzzy Time Series with and without Markov Chain using R Studio. The result showed that Fuzzy Time Series with and without Markov Chain method performs better in the forecasting ability in short-term period prediction and the values of Mean Absolute Percentage Error (MAPE) and Mean Square Error (MSE) tends to be smaller than the Fuzzy Time Series without Markov Chain. It can be concluded the number of exported can be used Fuzzy time series.

Short-Term Forecast of Blast Furnace Gas Production Amount Based on Grey RBF Neural Network

2015 ◽  
Vol 713-715 ◽  
pp. 1907-1913 ◽  
Author(s):  
Zhi Min Lv ◽  
Zhao Wang ◽  
Zi Yang Wang
Keyword(s):  
Neural Network ◽  
Blast Furnace ◽  
Dynamic Scheduling ◽  
Rbf Neural Network ◽  
Gas Production ◽  
Grey Theory ◽  
Short Term ◽  
Short Term Forecast ◽  
Blast Furnace Gas ◽  
Term Forecast

Dynamic optimization scheduling of the gas in iron and steel enterprises has great significance to reduce gas emission and the short-term forecast is the premise to realize the energy dynamic scheduling. Based on the characteristics that the influencing factors of blast furnace gas amount are complex and difficult to collect, a grey radial basis function (RBF) neural network forecast model is proposed to predict the gas amount for blast furnace in this paper. Combining grey theory, which is used to preprocess the historical data and obtain abundant information, with RBF neural network makes the effective trend forecast in the next 30 minutes come true. The model proposed in this paper is proved to be more accurate according to control experiments against the grey BP neural network.

scholarly journals Principles of organizing earthquake forecasting based on multiparameter sensor-WEB monitoring data

2020 ◽  
Vol 196 ◽  
pp. 03004
Author(s):  
Sergey Pulinets ◽  
Dimitar Ouzounov ◽  
Dmitry Davidenko ◽  
Pavel Budnikov
Keyword(s):  
Data Processing ◽  
Earthquake Forecasting ◽  
Time Interval ◽  
Satellite Measurements ◽  
Short Term ◽  
Term Forecast ◽  
Kamchatka Region ◽  
Multiparameter Monitoring ◽  
Intelligent Data Processing ◽  
Processing Service

The paper describes an approach that allows, basing on the data of multiparameter monitoring of atmospheric and ionospheric parameters and using ground-based and satellite measurements, to select from the data stream a time interval indicating the beginning of the final stage of earthquake preparation, and finally using intelligent data processing to carry out a short-term forecast for a time interval of 2 weeks to 1 day before the main shock. Based on the physical model of the lithosphere-atmospheric-ionospheric coupling, the precursors are selected, the ensemble of which is observed only during the precursory periods, and their identification is based on morphological features determined by the physical mechanism of their generation, and not on amplitude selection based on statistical data processing. Basing on the developed maquette of the automatic processing service, the possibility of real-time monitoring of the situation in a seismically active region will be demonstrated using the territory of the Kamchatka region and the Kuril Islands.

scholarly journals Real Time Short-term Forecasting Method of Remaining Parking Space in Urban Parking Guidance Systems

2018 ◽  
Vol 30 (2) ◽  
pp. 173-185 ◽  
Author(s):  
Xiaobo Zhu ◽  
Jianhua Guo ◽  
Wei Huang ◽  
Fengquan Yu ◽  
Byungkyu Brian Park
Keyword(s):  
Neural Network ◽  
Markov Chain ◽  
Real Time ◽  
Parking Space ◽  
Short Term ◽  
Markov Chain Method ◽  
Proposed Model ◽  
Forecasting Method ◽  
Short Term Forecasting ◽  
Urban Parking

Short-term forecasting of the remaining parking space is important for urban parking guidance systems (PGS). The previous methods like polynomial equations and neural network methods are difficult to be applied in practice because of low accuracy or lengthy initial training time which is unfavourable if real-time training is carried out on adapting to changing traffic conditions. To forecast the remaining parking space in real-time with higher accuracy and improve the performances of PGS, this study develops an online forecasting model based on a time series method. By analysing the characteristics of data collected in Nanjing, China, an autoregressive integrated moving average (ARIMA) model has been established and a real-time forecasting procedure developed. The performance of this proposed model has been further analysed and compared with the performances of a neural network method and the Markov chain method. The results indicate that the mean error of the proposed model is about 2 vehicles per 15 minutes, which can meet the requirements for general PGS. Furthermore, this method outperforms the neural network model and the Markov chain method both in individual and collective error analysis. In summary, the proposed online forecasting method appears to be promising for forecasting the remaining parking space in supporting the PGS.

Sign in / Sign up

Export Citation Format

Share Document