Mine fire evaluation system based on artificial intelligence algorithm

• 1、College of Safety Science and Engineering, Liaoning Technical University, Xingcheng 125100, China
• 2、School of Electrical Control and Engineering, Liaoning Technical University, Xingcheng 125100, China

Abstract：On the basis of systematic analysis of the occurrence factors of mine fire, a set of risk factors was established by using the analytic hierarchy process (AHP), which included three secondary factors: fire source factor, environment and its medium factor, and prevention and control technical measures. The three secondary factors were further divided into nine tertiary factors, and then constructed by three calculation weight methods and consistency test. In order to study the feedback of CO, CO2, CH4, C2H6, C2H4,N2 and other gases on temperature in the process of coal oxidation and heating up, and to predict the temperature of spontaneous combustion of coal based on the data of each gas, the gas component analysis method is adopted and the BP neural network prediction model is established. The concentration of each gas is taken as the input layer in the model, and the output layer is the temperature of coal seam. Finally, the purpose of temperature prediction is achieved by combining MATLAB. Finding the shortest path in the mine fire system is a basic problem of mine fire rescue. The application of ant colony algorithm and MATLAB can solve the shortest path to complete each node, providing theoretical support and practical execution route for rescue, thereby reducing the number of casualties in the rescue process

Keywords： Mine fire Analytic hierarchy process BP neural network Temperature prediction Ant colony algorithm Shortest path