摘要
为了预测倾斜厚大矿体自然崩落过程的岩层移动范围,选取矿体上盘围岩岩性、下盘围岩岩性、上盘围岩构造特征、下盘围岩构造特征、开采深度、开采厚度及矿体倾角为影响因素,以人工神经网络中的误差反向传播网络模型推导上盘、下盘岩层移动角。结果表明,BP神经网络预测Borska Reka厚大矿体超大规模自然崩落过程上盘岩层移动角为65.1°,下盘岩层移动角为72.2°,预测精度高,预测结果为合理布置地表工业场地及地表沉降监测提供了理论依据。
To predict the strata displacement range of inclined orebody during block caving,seven parameters were chosen as influential factors to evaluate the hanging wall and footwall strata displacements with error back propagation networks in artificial neural network.These parameters included hanging wall strata property,footwall strata property,hanging wall tectonic characteristics,footwall tectonic characteristics,mining depth,mining thickness and orebody inclination.Results show that the strata displacement angles at hanging wall and footwall are respectively 65.1°and 72.2°in massive block caving of thick Borska Reka orebody with BP neural network prediction.The prediction in high in accuracy and thus provides reasonable reference for arrangements of surface industrial sites and subsidence monitoring.
作者
李文东
简锡明
詹术霖
黄明清
Li Wendong;Jian Ximing;Zhan Shumin;Huang Mingqing(Zijin Mining Group Co.,Ltd.,Xiamen 361008,Fujian;Zijin School of Geology and Mining,Fuzhou University,Fuzhou 350108,Fujian)
出处
《福建冶金》
2023年第1期15-19,共5页
Fujian Metallurgy
基金
福建省自然科学基金(2019J05039)。
关键词
自然崩落
BP神经网络
岩层移动角
预测精度
block caving
BP neural network
strata displacement angle
prediction accuracy
