基于GA-BP神经网络的煤层底板突水量等级预测  

Prediction of Water Inrush Grade of Coal Seam Floor Based on GA-BP Neural Network

作　　者：刘艳冬 卢兰萍[1] 刘林林 王铁计 靳子栋 李大屯 LIU Yandong;LU Lanping;LIU Linlin;WANG Tieji;JIN Zidong;LI Datun(School of Civil Engineering,Hebei University of Engineering,Handan,Hebei 056038;Jizhong Energy Fengfeng Group,Handan,Hebei 056038;Handan Baofeng Co.,Ltd.Jiulong Mine,Handan,Hebei 056200)

机构地区：[1]河北工程大学土木工程学院,河北邯郸056038 [2]冀中能源峰峰集团,河北邯郸056038 [3]邯郸市宝峰有限公司九龙矿,河北邯郸056200

出　　处：《中国煤炭地质》2023年第8期32-37,共6页Coal Geology of China

基　　金：国家自然科学基金资助项目(41902254);河北省自然科学基金生态智慧矿山联合基金资助项目(D2020402013)。

摘　　要：BP神经网络虽然具备了解决非线性问题的能力,不过依然存在收敛速度慢,易陷入局部最优等问题,为了提高煤层底板突水量等级预测的准确性,提出一种新的方法,使用具有较强全局搜索能力的遗传算法来优化权值和阈值取代BP神经网络中随机初始的权值和阈值。以九龙矿区煤层底板突水实测资料为基础,建立了采用遗传算法优化的BP神经网络(GA-BP神经网络)预测煤层底板突水量等级模型。结果表明:该预测模型相对于BP神经网络模型预测性能更优,预测准确率提高了11%。Although BP neural network has the ability to solve nonlinear problems,it still meets the problems such as slow convergence speed and local optima.In order to improve the accuracy of coal seam floor water outburst grade prediction,a new method is proposed,that is,genetic algorithm with strong global search ability is used to optimize the weight.The threshold replaces the random initial weight threshold in BP neural network.Based on the measured data of coal seam floor water inrush in Jiulong Mine area,the BP neural network(GA-BP neural network)optimized by genetic algorithm is established to predict the water inrush grade of coal seam floor.The results show that the prediction performance of this model is better than that of BP neural network model,and the prediction accuracy is increased by 11%.

关 键 词：遗传算法优化 GA-BP神经网络 煤层底板 突水量等级 

分 类 号：TD742[矿业工程—矿井通风与安全]