新冒落带高度算法FOA-SVM预计模型  

New Roof Falling Highly Algorithm FOA-SVM Prediction Model

作　　者：尹志辉 孙邦达 

机构地区：[1]沈煤集团沈焦公司西马煤矿,辽宁辽阳111000 [2]浙江工商大学统计与数学学院,浙江杭州310018

出　　处：《煤炭与化工》2014年第6期10-13,共4页Coal and Chemical Industry

摘　　要：针对顶板冒落带高度问题提出新的预计模型,通过搜集众多矿井的实测数据,在支持向量机理论基础上建立预计模型。采用果蝇优化算法对预计模型进行优化训练,建立FOA-SVM预计模型,利用实测数据对模型的预计结果进行检验,预计结果较为准确,比PSO-SVM模型和GA-SVM模型结果稳定性好计算精度高。Based on the roof fall height problem, the new model was put forward, by collecting numerous experimental data of coal mine, based on support vector machine （SVM） theory, the expoeted model was settled. Using drosophila optimization algorithm optimized the new model training, the FOA - the SVM prediction model was established, using the measured data, the expected results of model test was expected more accurately, which had better stability calculation and more accuracy than PSO - SVM model and GA - SVM model results.

关 键 词：冒落带高度 支持向量机SVM 果蝇算法FOA 模型优化 

分 类 号：TD936[矿业工程—选矿]