基于CSAO-BP神经网络的山区开采沉陷预计  

Prediction of Mining Subsidence in Mountain Area Based on CSAO-BP Neural Network

作　　者：王雪英 张明媚 禹信 WANG Xueying;ZHANG Mingmei;YU Xin(Shanxi Institute of Energy,Jinzhong 030600,China)

机构地区：[1]山西能源学院,山西晋中030600

出　　处：《煤炭技术》2023年第6期41-44,共4页Coal Technology

基　　金：山西省高等学校科技创新项目(2021L596)。

摘　　要：针对山区开采沉陷预计较复杂的情况,提出一种基于混沌螺旋天鹰优化算法优化BP神经网络的CSAO-BP开采沉陷预计模型。首先介绍基本的天鹰算法,然后将螺旋觅食策略、Tent混沌映射和边界控制策略结合对AO进行优化,进而改进传统BP模型,建立AO-BP和CSAO-BP预计模型;然后,选取数值各异的32组数据作为样本集,并分别用传统BP模型、AO-BP模型、CSAOBP模型进行测试;结果表明,CSAO-BP模型精度明显优于AO-BP模型和传统BP模型,在山区开采沉陷领域具有可行性和实用性。In view of the complexity of mining subsidence prediction in mountainous areas,proposes a CSAO-BP mining subsidence prediction model based on BP neural network optimized by chaotic spiral skyhawk optimization algorithm.Firstly,the basic skyhawk algorithm is introduced,and then the spiral foraging strategy,Tent chaotic mapping and boundary control strategy are combined to optimize AO,and then the traditional BP model is improved to establish AO-BP and CSAO-BP prediction models;then,32 groups of data with different values are selected as the sample set and tested with traditional BP model,AO-BP model and CSAO-BP model respectively;the results show that the precision of CSAO-BP model is obviously superior to AO-BP model and traditional BP model,and it is feasible and practical in the field of mining subsidence in mountain areas.

关 键 词：开采沉陷 天鹰优化算法 混沌螺旋天鹰算法 神经网络 CSAO-BP神经网络 

分 类 号：TD325[矿业工程—矿井建设]