基于神经网络的采空区自燃预测

Gob spontaneous combustion prediction based on neural network

针对矿井采空区自燃情况难以监测的状况,提出将遗传算法与神经网络应用于采空区自燃预测的方法。该算法利用遗传算法的全局搜索能力强的特点去优化神经网络,由神经网络构成的推理系统去预测采空区的自燃报警。仿真结果表明:当瓦斯体积分数和温度处于危险等级时,可以准确地预测出危险报警,验证了该方法的有效性和可靠性。

Aiming at the hard condition of mine gob spontaneous combustion prediction,a scheme that applies GA and neural network in gob spontaneous combustion prediction is proposed.This algorithm optimizes neural network by using GA＇s strong global search capability,and predict the gob spontaneous combustion alarm by the inference system of neural network.The simulation results show that when the gas volume fraction and temperature are in danger level,this scheme can predict the danger and alarm accurately,it demonstrates the scheme＇s effectiveness and reliability.