Study on condition analysis and temperature prediction of coal spontaneous combustion based on improved genetic algorithm

Abstract

For major coal mine spontaneous combustion caused by major disasters every year, various scholars have studied and analyzed the conditions of coal mine spontaneous combustion and predicted the coal temperature. Coal mine spontaneous combustion is an inevitable disaster, but studying the coal mine surface covering to damp coal mine spontaneous combustion can greatly reduce the occurrence of coal mine spontaneous combustion. We analyzed the oxygen absorption in the pores of coal and the tendency of its own water content for spontaneous combustion of coal. The model experiment was carried out, and the correlation between different gas concentrations and coal temperature produced during spontaneous combustion of mixed coal samples was analyzed. The coal temperature was predicted in coal mines with different water contents and oxygen absorption levels, the coal temperature was predicted four times, and the performance evaluation and comparison of the correct prediction of coal temperature under different algorithm models were carried out. Finally, in the contrast experiment, the curve of the fitting function and minimum error value is further compared, and it can be seen that the combined model of the genetic algorithm and neural network algorithm has more accurate prediction accuracy than the single model. Through the analysis of the coal spontaneous combustion phenomenon and the study of anti-coal spontaneous combustion devices and the oxidation process, we have reduced the risk of spontaneous combustion in coal mining areas.