Research On The Application Of Gas Emission Prediction In Mine Excavation Faces

According to the 14 th Five Year Plan for National Economic and Social Development of the People’s Republic of China and the Outline of Long Range Objectives for 2035,"safety risk monitoring and early warning" is a key planning and construction content.One of the core contents of "safety risk monitoring and early warning" is the prediction and early warning of gas in mines.Therefore,in order to prevent and control gas explosions and ensure mine safety,it is necessary to construct a new prediction method for gas emission to improve the prediction effect and ensure the safety of workers’ lives and property.In response to the above issues,the paper mainly completed the following three tasks:(1)Analyze the factors affecting the amount of gas emitted from the mine.Based on the comprehensive characteristics of gas and various geological factors,a factor matrix for gas emission is constructed for geological theoretical analysis and curve fitting.Further study the positive and negative correlations between gas emission and other gas influencing factors based on curve fitting data,and use kernel principal component analysis to extract features from the factor matrix and construct input and output gas emission data matrices.(2)The prediction algorithm of mine gas emission is constructed based on the peacock optimization algorithm-long short-term memory networks.The three-layer network architecture of the long short-term memory networks is optimized according to the energy search behavior,adaptive approach behavior,and cub adaptive search behavior of the peacock optimization algorithm.The improved long short-term memory networks is used to predict the long-term gas emission.(3)Construct a sparrow search algorithm-extreme learning machine for predicting mine gas emission.The chimeric sparrow search algorithm-extreme learning machine adds chaotic mapping to the sparrow search algorithm to expand its search space,and improves parameters such as the population size and learning factor of the sparrow search algorithm.Embed the improved sparrow search algorithm into the extreme learning machine to optimize its weights and nodes.Apply the improved extreme learning machine to short-term gas emission prediction.