The use of the neuro-fuzzy model to predict the methane hazard during the underground coal mining production process

Abstract

Coal mining production is marked with a number of different types of hazards, of which the most dangerous are natural hazards resulting from disturbances in the rock mass balance due to mining operations. One of the most frequent and dangerous to the continuity, effectiveness and safety of this process is a methane hazard. It is cause by methane, a gas naturally occurring in coal beds. In order to limit this threat, the article presents a developed methodology of its diagnosis and prognosis together with an example of its practical application. The methodology is based on the neural-fuzzy model, which, using the results of measurements of real ventilation parameters, enables a diagnosis (fuzzy model) and forecast (neural-fuzzy model) the degree of methane hazard in the area of exploitation. The measure of this hazard is the value of the methane hazard index (MHI), which takes into account the relation between the absolute and criterial methane-bearing capacity in the examined region. The developed models and methodology enable the determination of short-term methane susceptibility prognosis, which provides an opportunity for current and effective methane susceptibility control in a given mining exploitation area. This in turn, by taking appropriate actions (e.g. change of ventilation parameters), makes it possible to improve safety and thus the effectiveness of the whole mining production process. The presented example of using this methodology in real conditions confirms its effectiveness. This mainly concerns the pace and reliability of the obtained results and the possibility of continuous learning of the developed system based on new data. The paper is an example of the original application of advanced methods of modeling of physical phenomena to solve a practical problem of improving the safety of the production process.