工作面多变量瓦斯体积分数时间序列预测模型

被引：5

作者：

董丁稳

李树刚

常心坦

林海飞

机构：

[1] 西安科技大学能源学院

来源：

采矿与安全工程学报 | 2012年 / 29卷 / 01期

基金：

高等学校博士学科点专项科研基金;

关键词：

贝叶斯网络; 相空间重构; 高斯过程; 多变量时间序列; 区间预测; 瓦斯体积分数;

D O I：

暂无

中图分类号：

TD712 [矿井瓦斯];

学科分类号：

081903 ;

摘要：

为了有效分析煤矿瓦斯监测数据以实现较准确的工作面瓦斯体积分数预测,基于贝叶斯网络方法、混沌相空间重构技术与高斯过程回归模型,研究了瓦斯体积分数时间序列分析与预测的方法。应用贝叶斯网络方法提取与工作面瓦斯体积分数时间序列有较强关联特征的样本数据集,构建了多变量瓦斯体积分数时间序列预测模型;采用混沌相空间重构技术来实现多变量瓦斯体积分数时间序列样本空间重构;应用高斯过程回归模型进行工作面多变量瓦斯体积分数预测,以预测值及其置信区间来表达对工作面未来瓦斯体积分数动态变化的预测。实例分析表明:应用该方法得到的预测结果,其预测精度较单变量瓦斯体积分数时间序列预测有较大提升,并且预测区间在相同置信水平下达到了最优,能够较好的反映工作面瓦斯体积分数的动态变化状况。

引用

页码：135 / 139

页数：5

共 14 条

[1]

On-line least squares support vector machine algorithm in gas prediction[J]. ZHAO Xiao-hu,WANG Gang,ZHAO Ke-ke,TAN De-jian School of Information & Electronic Engineering,China University of Mining & Technology,Xuzhou,Jiangsu 221008,China. Mining Science and Technology. 2009(02)

[2]

Advances in gaussian processes. RASMUSSEN C E. Proceedings of the Neural Information ProcessingSystems Conference . 2006

[3]

Gaussian process approachfor modeling of nonlinear systems. GREGOR G,GORDON L. EngineeringApplications of Artificial Intelligence . 2009

[4]

Introduction to bayesian networks. ZHANG N L,GUO H. . 2006

[5]

Regional predictionof coal and gas outburst hazard based on multi-factorpattern recognition. SONG Wei-hua,ZHANG Hong-wei. Procedia Earth and PlanetaryScience . 2009

[6]

A new kernel-basedapproach for linear system identification. GIANLUIGI P,GIUSEPPE D N. Automatica . 2010

[7]

Application of greyrelation clustering and CGNN in gas concentrationprediction in top corner of coal mine. QU Zhi-ming,LIANG Xiao-ying. 2009International Conference on Computational Intelligenceand Natural Computing . 2009

[8]

A constructiveneural network learning method based on quotient spaceand its application in coal mine gas prediction. LIU Yu-jun,ZHANG Yue-qin,ZHU Yu. 2010International Conference on Intelligent Computing andCognitive Informatics . 2010

[9]

Coalmine gas concentration forecasting based on chaotictheory and neural network model. ZHAO Jin-xian,JIN Hong-zhang,YU Guang-hua. Proceedings of theThird International Conference on Information andComputing Science . 2010

[10]

Differential phase space reconstructedfor chaotic time series. XU Peng-cheng. Applied MathematicalModeling . 2009

← 1 2 →