Forecasting the aerogas state of mine atmosphere with application of artificial neural networks
/ 0
- Details
- Category: Information technologies, systems analysis and administration
- Last Updated on 07 January 2019
- Published on 26 December 2018
- Hits: 3194
Authors:
H.H.Shvachych, Dr. Sc. (Tech.), Prof., orcid.org/0000-0002-9439-5511, National Metallurgical Academy of Ukraine, Dnipro, Ukraine, e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
O.V.Ivaschenko, orcid.org/0000-0003-4394-6907, National Metallurgical Academy of Ukraine, Dnipro, Ukraine, e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
Ye.Ye.Fedorov, Dr. Sc. (Tech.), Assoc. Prof., orcid.org/0000-0003-3841-7373, Donetsk National Technical University, Pokrovsk, Donetsk region, Ukraine
Yu.L.Dikova, orcid.org/0000-0002-8196-0817, Donetsk National Technical University, Pokrovsk, Donetsk region, Ukraine
Abstract:
Purpose. Development of a method for forecasting a highly dynamic process of changing the concentration of explosive gases in a mine using artificial neural networks that allow considering external factors. Nowadays, increase in industrial safety in the mining industry is solved at the expense of introduction of modern computer systems of the aerogas control at the enterprises; the systems’ work is aimed at the forecast of parameters of the environment condition. In this regard, the introduction of computer systems based on the use of artificial neural networks allows providing valid recommendations for the adoption of optimal technological and management solutions.
Methodology. Forecasting the aerogas state of the mine atmosphere is based on the use of artificial neural networks, autoregressive models and meta-heuristics.
Findings. Nowadays, the main measures for the forecast of methane concentration are aimed at finding the dynamics regularities of gas concentration, which served as the basis for forecasting gas-dynamic phenomena. The main results were obtained in studies of the dynamics of methane concentration using telemetric monitoring devices. This approach is based on the use of linear models. The proposed use of non-linear forecast models based on artificial neural networks provides more accurate forecasts comparing to widely used linear models.
Originality. In the framework of forecasting the air-gas state of the mine atmosphere, artificial neural networks and autoregressive models were further developed. At the same time, the autoregressive forecast model was improved by adding the exogenous factors to its structure, which are the measured dynamic parameters of the aerogas state of mine developments. To adapt the model, the meta-heuristic algorithm is improved.
Practical value. The results of the experiments showed that the proposed artificial neural network proved to be more effective for solving the problem of forecasting the aerogas state of the mine atmosphere in comparison with the existing approach. Numerical studies have shown that the proposed model allows increasing accuracy of the forecast by 10 % in comparison with the widely used gradient methods. The positive economic effect from introduction of the proposed approach is to reduce the likelihood of emergencies, which leads to a reduction in financial costs aimed at eliminating the consequences of accidents.
References.
1. Methane control [online]. Available at: <http://vovremya.info/art/1196506356.html> [Accessed 11 November 2017].
2. Vovna, O. V. and Zori, A. A., 2015. Subsystem of limits control of explosion of a mine atmosphere for a system of aerogas protection of coal mines. Visn. NTU “KhPI”. Scientific proceedings. Power and conversion technology. Kharkiv, 19 (1128), pp. 79‒88.
3. Vovna, O. V. and Zori, A. A., 2014. Optical concentration meter with hardware and software compensation of temperature drift. Acad. Papers od Donetsk Nat. Tech. University. Computer science and automatization, 1(26), pp. 178‒188.
4. Regression Rule Learning for Methane Forecasting in Coal Mines [online]. Available at: <http://link.springer.com/chapter/10.1007%2F978-3-319-18422-7_44#page-1>[Accessed 20 September 2017].
5. Fedorov, E. E. and Dikova, Yu. L., 2016. Multiagent system for forecasting the state of the miner atmosphere. In: SIMULATION-2016 ‒ Kyiv, conference proceedings, pp. 61‒65. Available at: <http://ea.donntu.edu.ua:8080/bitstream/123456789/28134/1/%D0%BC%D1%83%D0%BB%D1%8C%D1%82%D0%B8%D0%B0%D0%B3%D0%B5%D0%BD%D1%82%D0%B0%D1%8F%20%D1%81%D0%B8%D1%81%D1%82%D0%B5%D0%BC%D0%B0%20%D0%BF%D1%80%D0%BE%D0%B3%D0%BD%D0%BE%D0%B7%D0%B0%20%D1%81%D0%BE%D1%81%D1%82%D0%BE%D1%8F%D0%BD%D0%B8%D1%8F%20%D1%80%D1%83%D0%B4%D0%BD%D0%B8%D1%87%D0%BD%D0%BE%D0%B9%20%D0%B0%D1%82%D0%BC%D0%BE%D1%81%D1%84%D0%B5%D1%80%D1%8B%20.pdf> [Accessed 05 September 2017].
6. Integrated security systems [online]. Available at: <http://itras.com.ua/> [Accessed 5 December 2017].
Archive