基于GWO-SVR混合骨料膏体流变参数预测模型  

Prediction Model of Rheological Parameters of Mixed Aggregate Paste based on GWO-SVR

作　　者：李聪 乔登攀[1] 杨天雨[1] 王俊[1] 陈杰 张希 LI Cong;QIAO Dengpan;YANG Tianyu;WANG Jun;CHEN Jie;ZHANG Xi(School of Land Resources Engineering,Kunming University of Science and Technology,Kunming 650093,China;Datun Tin Mine,Yunnan Tin Co.,Ltd.,Gejiu 661000,China;Liangshan Mining Co.,Ltd.,Liangshan 615141,China)

机构地区：[1]昆明理工大学国土资源工程学院,昆明650093 [2]云南锡业股份有限公司大屯锡矿,云南个旧661000 [3]凉山矿业股份有限公司,四川凉山615141

出　　处：《有色金属工程》2023年第2期127-138,共12页Nonferrous Metals Engineering

基　　金：云南省基础研究专项-青年项目(202101AU070022);昆明理工大学人培基金(KKZ3202021040)。

摘　　要：膏体的流变性能对充填管路合理设计和管道输送阻力准确分析有重要影响。为了对膏体流变参数进行准确预测,利用灰狼算法(GWO)对支持向量回归(SVR)进行参数寻优,建立了以混合骨料掺和比、膏体质量分数和水灰比为输入参数的SVR和GWO-SVR两种屈服应力和黏度预测模型,并且对两种模型进行了对比分析。结果表明:GWO-SVR模型与传统SVR模型相比,屈服应力和黏度预测模型的决定系数R^(2)分别提高了5.8%和4.05%,并且GWO-SVR模型预测结果误差更小,屈服应力和黏度预测值相比试验值,其最大误差分别为5.63 Pa和0.00685 Pa·s,表明GWO-SVR模型有着更高的精度,可对混合骨料膏体屈服应力和黏度进行有效预测。The rheological property of paste has an important influence on the rational design of filling pipeline and the accurate analysis of pipeline transportation resistance.In order to accurately predict the rheological parameters of the paste,the grey wolf algorithm(GWO)is used to optimize the parameters of support vector regression(SVR),and two prediction models of yield stress and viscosity,SVR and GWO-SVR,which take the mixing ratio,mass fraction and water cement ratio of mixed aggregate as input parameters.The two models are compared and analyzed.The results show that,compared with the traditional SVR model,the determination coefficient R^(2) of the yield stress and viscosity prediction model of GWO-SVR model is increased by 5.8%and 4.05%respectively,and the error of the prediction result of GWO-SVR model is smaller.The maximum error of the yield stress and viscosity prediction value is 5.63 Pa and 0.00685 Pa·s respectively compared with the test value,which indicates that GWO-SVR model has higher accuracy and can effectively predict the yield stress and viscosity of mixed aggregate paste.

关 键 词：混合骨料 充填膏体 流变参数预测 灰狼算法 支持向量回归 

分 类 号：TD853.34[矿业工程—金属矿开采]