基于MPSO-SVM巷道围岩松动圈预测研究  

Prediction model of loosening zones around roadway based on MPSO-SVM

作　　者：朱志洁[1] 张宏伟[1] 陈蓥[1] 

机构地区：[1]辽宁工程技术大学矿业学院,辽宁阜新123000

出　　处：《计算机工程与应用》2014年第12期1-5,共5页Computer Engineering and Applications

基　　金：辽宁省教育厅基金资助项目(No.L2012106)

摘　　要：针对目前巷道围岩松动圈确定方法的种种缺陷,提出了一种新的预测方法,采用改进的粒子群算法(MPSO)优化支持向量机(SVM)对巷道围岩松动圈进行预测。在标准PSO中引入压缩因子,实现了算法全局搜索和局部寻优的有效平衡;应用MPSO对SVM的参数C和g进行优化,建立MPSO-SVM回归预测模型;将该预测模型应用于巷道围岩松动圈的预测,将预测性能与PSO-SVM、GA(遗传算法)-SVM、GSM(网格搜索)-SVM模型、BP神经网络进行对比分析。结果表明:该模型具有较强的泛化能力,较高的预测精度,可以对围岩松动圈厚度进行有效预测。A new method using Support Vector Machine（SVM）which is optimized by Modified Particle Swarm Optimi-zation（MPSO）to predict loosening zones around roadway is proposed, avoiding shortcomings of current determining methods. Compression factor is introduced to standard PSO. Effective balance between global search and local optimiza-tion is achieved. MPSO is applied to optimizing the SVM parameters C and g, and MPSO-SVM regression prediction model is established. The model is applied to predicting loosening zones around roadway. The prediction result is com-pared with PSO-SVM, GA（Genetic Algorithm）-SVM, GSM（Grid Search Method）-SVM, BP neural network. The results show that the model has better generalization performance and higher prediction accuracy. It can effectively predict loos-ening zones around roadway.

关 键 词：粒子群算法 支持向量机 围岩松动圈 仿真预测 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] X936[自动化与计算机技术—控制科学与工程]