基于神经网络的上隅角CO涌出影响因素分析

Analysis of Influencing Factors of Coal Gangue Emissions at Upper Corner based on Neural Network

针对回采巷道上隅角CO涌出量超标的问题,对影响CO涌出的因素进行了分析,选取了漏风、回风、周期压力、氧气浓度、推进量五个因素为研究对象,根据实测数据的仿真确定了隐含层神经元数量和训练函数的选取,确定了神经网络模型。采用网络权重矩阵的权重贡献率公式计算得出了各影响因素的重要排序,为井下CO控制提供了理论依据。

Aiming at the problem of excessive CO emission at the upper corner of the roadway, the factors infuencing CO emission in coal mine are analyzed. Five factors including air leakage, return air, periodic pressure, oxygen concentration and propulsive quantity were selected as the research objects. The neural network prediction model determines the number of neurons in the hidden layer and the selection of training functions based on the simulation of measured data, and it also determines the neural network model. By using the weighted contribution formula of network weight matrix, the important ranking of each infuencing factor is obtained, which provides a theoretical basis for underground CO control.