硫化矿石自燃倾向性等级划分的支持向量机模型及应用
|
摘要
为准确判定硫化矿石的自燃倾向性等级,提出一种硫化矿石自燃倾向性等级划分预测模型——支持向量机(Support Vector Machine,SVM)模型。结合已有的研究成果,选取表征硫化矿石自燃倾向性本质特性的自热点温度、室内低温氧化质量增加率和自燃点温度3项指标作为硫化矿自燃倾向性等级划分的基本判别因子。使用典型高硫矿山的18组矿样的实测数据作为训练样本,6组硫化矿井矿样的自燃倾向性作为预测样本。分别采用网格参数寻优、遗传算法(Genetic Algorithm,GA)参数寻优、粒子群(Particle Swarm Optimization,PSO)寻优算法来计算SVM模型的参数,分析比较三种算法得到的预测结果,确定最适合硫化矿石自燃倾向性等级划分的SVM回归模型的参数寻优算法。研究结果表明,网格寻优算法、GA算法取得了良好的预测效果,预测正确率为100%。因此,支持向量机模型可以用于指导高硫矿山矿石自燃倾向性等级的划分。
Support vector machine( SVM) model is proposed to determine the spontaneous combustion tendency of sulfide ores. Based on the existing research results,the characteristics of spontaneous combustion tendency of sulfide ores were selected as the hot spot temperature,the increase rate of the indoor low temperature oxidation and the temperature of the ignition point as the basic criterion for the classification of spontaneous combustion tendency of sulfide ores. Using measured data of typical high sulfur mine 18 gro-ups of samples as training samples,6 groups of sulfide mine ore spontaneous combustion tendency as prediction samples. The parameters of SVM model are calculated by using rid search algorithm,genetic algor-ithm( GA) and particle swarm optimization( PSO). The parameters of the three algorithms are analyzed and compared. The results show that the grid search algorithm and the GA algorithm have achieved good results,and the prediction accuracy is 100%. Therefore,support vector machine model can be used to guide the classification of ore spontaneous combustion tendency in high sulfur mine.
Support vector machine( SVM) model is proposed to determine the spontaneous combustion tendency of sulfide ores. Based on the existing research results,the characteristics of spontaneous combustion tendency of sulfide ores were selected as the hot spot temperature,the increase rate of the indoor low temperature oxidation and the temperature of the ignition point as the basic criterion for the classification of spontaneous combustion tendency of sulfide ores. Using measured data of typical high sulfur mine 18 gro-ups of samples as training samples,6 groups of sulfide mine ore spontaneous combustion tendency as prediction samples. The parameters of SVM model are calculated by using rid search algorithm,genetic algor-ithm( GA) and particle swarm optimization( PSO). The parameters of the three algorithms are analyzed and compared. The results show that the grid search algorithm and the GA algorithm have achieved good results,and the prediction accuracy is 100%. Therefore,support vector machine model can be used to guide the classification of ore spontaneous combustion tendency in high sulfur mine.
引文
[1]阳富强,刘广宁,郭乐乐.硫化矿石自燃倾向性等级划分的GA-BP神经网络模型及应用[J].自然灾害学,2015,24(4):227-232.
[2]胡文志,谢振华,赵军.硫化矿石自燃倾向性鉴定技术[J].金属矿山,2011,35(6):146-148.
[3]谢正文,吴超,李孜军,等.基于信息熵和集对分析理论的硫化矿石自燃倾向性判定[J].中南大学学报(自然科学版),2012,43(5):1 858-1 863.
[4]阳富强,陈伯辉.硫化矿石自燃倾向性评价的属性区间识别模型[J].中国安全科学学报,2012,22(1):70-75.
[5]赵军,张兴凯,王云海,等.基于统计回归模型的硫化矿石自燃倾向性鉴定指标研究[J].中国安全生产科学技术,2010,6(4):20-22.
[6]胡汉华,刘征,李孜军,等.硫化矿石自燃倾向性等级分类的Fisher判别分析法[J].煤炭学报,2010,35(10):1 674-1 679.
[7]阳富强,吴超,李孜军,等.基于距离判别分析理论的硫化矿石自燃倾向性等级划分[J].煤炭学,2010,35(12):2 111-2 115.
[8]阳富强,吴超,李孜军.硫化矿石自燃倾向性综合判定的物元模型及其应用[J].中南大学学报(自然科学版),2011,42(11):3 459-3 464.
[9]赵军,张兴凯,王云海.硫化矿石自燃倾向性鉴定技术研究[J].中国安全生产科学技术,2009,5(6):105-109.
[10]余志雄,周创兵,李俊平,等.基于ν-SVR算法的边坡稳定性预测[J].岩石力学与工程学报,2005,24(14):2 468-2 475.
[11]罗战友,杨晓军,龚晓南.基于支持向量机的边坡稳定性预测模型[J].岩石力学与工程学报,2005,24(1):144-148.
[12]陈祖云,张桂珍,邬长福,等.基于支持向量机的边坡稳定性预测研究[J].中国安全生产科学技术,2009,5(4):101-105.
[13]曹庆奎,王瑞.基于GA-LSSVR的露天矿边坡稳定性预测[J].煤矿安全,2014,45(2):200-203.
[2]胡文志,谢振华,赵军.硫化矿石自燃倾向性鉴定技术[J].金属矿山,2011,35(6):146-148.
[3]谢正文,吴超,李孜军,等.基于信息熵和集对分析理论的硫化矿石自燃倾向性判定[J].中南大学学报(自然科学版),2012,43(5):1 858-1 863.
[4]阳富强,陈伯辉.硫化矿石自燃倾向性评价的属性区间识别模型[J].中国安全科学学报,2012,22(1):70-75.
[5]赵军,张兴凯,王云海,等.基于统计回归模型的硫化矿石自燃倾向性鉴定指标研究[J].中国安全生产科学技术,2010,6(4):20-22.
[6]胡汉华,刘征,李孜军,等.硫化矿石自燃倾向性等级分类的Fisher判别分析法[J].煤炭学报,2010,35(10):1 674-1 679.
[7]阳富强,吴超,李孜军,等.基于距离判别分析理论的硫化矿石自燃倾向性等级划分[J].煤炭学,2010,35(12):2 111-2 115.
[8]阳富强,吴超,李孜军.硫化矿石自燃倾向性综合判定的物元模型及其应用[J].中南大学学报(自然科学版),2011,42(11):3 459-3 464.
[9]赵军,张兴凯,王云海.硫化矿石自燃倾向性鉴定技术研究[J].中国安全生产科学技术,2009,5(6):105-109.
[10]余志雄,周创兵,李俊平,等.基于ν-SVR算法的边坡稳定性预测[J].岩石力学与工程学报,2005,24(14):2 468-2 475.
[11]罗战友,杨晓军,龚晓南.基于支持向量机的边坡稳定性预测模型[J].岩石力学与工程学报,2005,24(1):144-148.
[12]陈祖云,张桂珍,邬长福,等.基于支持向量机的边坡稳定性预测研究[J].中国安全生产科学技术,2009,5(4):101-105.
[13]曹庆奎,王瑞.基于GA-LSSVR的露天矿边坡稳定性预测[J].煤矿安全,2014,45(2):200-203.