基于FOA-SVM的煤矿瓦斯爆炸风险模式识别
|
摘要
瓦斯爆炸的过程是一个综合各种因素在内的能量释放的过程,为了能够对瓦斯爆炸进行准确的辨识,文中提出了将主成分分析(Principal Componet Analysis,PAC)和支持向量机(Support Vector Machine,SVM)相结合进行瓦斯爆炸预测。瓦斯爆炸影响因素较多,首先通过PCA进行特征提取,降低特征向量的维数,去除参数间的相关性;然后通过果蝇优化算法(Fruit Fly Optimization Algorithm,FOA)对支持向量机进行全局寻优,避免了过学习的出现。将通过PCA提取的新特征值作为FOA-SVM模型输入,从而实现准确性高的瓦斯爆炸风险模式识别。通过实验仿真表明,文中提出的方法具有识别精度高、推广能力强同时模型简单的特点,对工程实践具有一定的指导作用。
Gas explosion is a releasing energy process combining with factors. In order to identify the risk of gas explosion accurately, this paper puts forward the combination of the principal componet analysis(PCA) and the support vector machine(SVM) for pattern recognition. Firstly, feature extraction is performed by PCA to reduce the dimension of the feature vector, then each parameter will be orthogonal; then through the fruit fly optimization algorithm(FOA), parameters of SVM are comprehensively optimized, which can avoid the emergence of over learning and less learning; the new eigenvalues extracted by PCA are the FOA-SVM model's inputs. The trained PCA and FOA-SVM model have the characterstics of high recognition rate and fast running for the risk identification of gas explosion.
Gas explosion is a releasing energy process combining with factors. In order to identify the risk of gas explosion accurately, this paper puts forward the combination of the principal componet analysis(PCA) and the support vector machine(SVM) for pattern recognition. Firstly, feature extraction is performed by PCA to reduce the dimension of the feature vector, then each parameter will be orthogonal; then through the fruit fly optimization algorithm(FOA), parameters of SVM are comprehensively optimized, which can avoid the emergence of over learning and less learning; the new eigenvalues extracted by PCA are the FOA-SVM model's inputs. The trained PCA and FOA-SVM model have the characterstics of high recognition rate and fast running for the risk identification of gas explosion.
引文
[1]李润求,施式亮,念其锋.近10年我国煤矿瓦斯灾害事故规律研究[J].中国安全科学学报,2011,21(9):144-151.Li R Q,Shi S L,Nian Q F.Research on Coalmine Gas Accident Rules in China in Recent Decade[J].China Safety Science Journal,2011,21(9):144-151.
[2]李凡修,梅平.煤矿安全评价的等效数值模型及应用[J].矿业安全与环保,2009,36(1):86-87.Li F X,Mei P.Equivalent Numerical Model and Application of Coal Mine Safety Evaluation[J].Mining Safety and Environmental Protection,2009,36(1):86-87.
[3]潘玉民,邓永红,张全柱,等.基于QPSO-RBF的瓦斯涌出量预测模型[J].中国安全科学学报,2012,22(12):29-34.Pan Y M,Deng Y H,Zhang Q Z,et al.Gas Emission Prediction Model Based on QPSO-RBF[J].China Safety Science Journal,2012,22(12):29-34.
[4]潘竞涛.工作面瓦斯涌出量逆向推算瓦斯含量梯度[J].辽宁工程技术大学学报(自然科学版),2012,31(6):818-821.Pan J T.Study on reverse projection of gas concentration gradient using face gas emission sub-source prediction method[J].Journal of Liaoning Technical University(Natural Science),2012,31(6):818-821.
[5]闫孝姮,陈伟华,付华.一种瓦斯涌出量系统预测辨识模型的研究[J].控制工程,2014,21(4):592-595.Yan X H,Chen W H,Fu H.Study on a New Predicted and Discernible Model of Gas Emission System[J].Control Engineering of China,2014,21(4):592-595.
[6]付华,王馨蕊,杨本臣,等.基于MPSO-CWLS-SVM的瓦斯涌出量预测[J].传感技术学报,2014,24(11):1568-1572.Fu H,Wang X R,Yang B C,et al.The Prediction of Mine-Gas Emission Based on MPSO-CWLS-SVM[J].CHINESE JOURNALOF SENSORS AND ACTUATORS,2014,24(11):1568-1572.
[7]曲方,安文超,李迎业,等.基于GRA-SVM方法的煤与瓦斯突出预测模型研究[J].中国煤炭,2012,38(11):102-106.Qu F,An W C,Li Y Y,et al.Research on prediction model for coal and gas outburst based on GRA-SVM[J].CHINA COAL,2012,38(11):102-106.
[8]张克.基于地震演模拟和SVM的煤与瓦斯突出危险区预测研究[D].中国矿业大学,2011,40-76.Zhang K.Research on Prediction of Coal and Gas Outburst Risk Area Based on Seismic Forward Modeling and SVM[D].China Mining University,2011,40-76.
[9]黄为勇,邵晓根,陈奎.一种采用CCPSO-SVM的煤与瓦斯突出预测方法[J].计算机科学,2012,39(11):216-225.Huang W Y,Shao X G,Chen K.Coal-and-Gas Outburst Forecast Using CCPSO and SVM[J].Computer Science,2012,39(11):216-225.
[10]刘俊娥,曾凡雷,郭章林.基于RS-SVM模型的煤与瓦斯突出多因素风险评价[J].中国安全科学学报,2011,21(7):21-25.Liu J E,Zeng F L,Guo Z L.Study on Multiple Factors Risk Evaluation of Coal and Gas Outburst Based on RS-SVM Model[J].China Safety Science Journal,2011,21(7):21-25.
[11]李润求,施式亮,念其锋.基于PSO-SVM煤矿瓦斯爆炸灾害风险模式识别[J].中国安全科学学报,2013,23(5):38-43.Li R Q,Shi S L,Nian Q F.Research on Pattern Recognition of Gas Explosion Disaster Risk in Coal Mines Based on PSO-SVM[J].China Safety Science Journal,2013,23(5):38-43.
[12]程东全,顾锋.基于可拓展模式识别煤与瓦斯突出危险性分析[J].安全与环境学报,2012,12(1):241-244.Chen D Q,Gu F.Risk analysis of coal and gas outburst based on extensible pattern recognition[J].Journal of Safety and Environment,2012,12(1):241-244.
[13]Hojjat Akhondi-Asl,James D.B.Nelson.M-Estimate robust PCAfor Seismic Noise Attenuation[C].IEEE International Conference on Image Processing,2016,1853-1857.
[14]Pan W T.A new fruit fly optimization algorithm:taking thefinancial distress model as an example[J].Knowledge-Based-Systems,2012,26:69-74.
[15]王雨虹,付华,张洋.基于KPCA和CIPSO-PNN的煤与瓦斯突出强度辨识模型[J].传感技术学报,2014,27(11):271-277.Wang Y H,Fu H,Zhang Y.The Identification Model of Coal and Gas OutburstIntensity Based on KPCA and CIPSO-PNN[J].CHINESE JOURNAL OF SENSORS AND ACTUATORS,2014,27(11):271-277.
[2]李凡修,梅平.煤矿安全评价的等效数值模型及应用[J].矿业安全与环保,2009,36(1):86-87.Li F X,Mei P.Equivalent Numerical Model and Application of Coal Mine Safety Evaluation[J].Mining Safety and Environmental Protection,2009,36(1):86-87.
[3]潘玉民,邓永红,张全柱,等.基于QPSO-RBF的瓦斯涌出量预测模型[J].中国安全科学学报,2012,22(12):29-34.Pan Y M,Deng Y H,Zhang Q Z,et al.Gas Emission Prediction Model Based on QPSO-RBF[J].China Safety Science Journal,2012,22(12):29-34.
[4]潘竞涛.工作面瓦斯涌出量逆向推算瓦斯含量梯度[J].辽宁工程技术大学学报(自然科学版),2012,31(6):818-821.Pan J T.Study on reverse projection of gas concentration gradient using face gas emission sub-source prediction method[J].Journal of Liaoning Technical University(Natural Science),2012,31(6):818-821.
[5]闫孝姮,陈伟华,付华.一种瓦斯涌出量系统预测辨识模型的研究[J].控制工程,2014,21(4):592-595.Yan X H,Chen W H,Fu H.Study on a New Predicted and Discernible Model of Gas Emission System[J].Control Engineering of China,2014,21(4):592-595.
[6]付华,王馨蕊,杨本臣,等.基于MPSO-CWLS-SVM的瓦斯涌出量预测[J].传感技术学报,2014,24(11):1568-1572.Fu H,Wang X R,Yang B C,et al.The Prediction of Mine-Gas Emission Based on MPSO-CWLS-SVM[J].CHINESE JOURNALOF SENSORS AND ACTUATORS,2014,24(11):1568-1572.
[7]曲方,安文超,李迎业,等.基于GRA-SVM方法的煤与瓦斯突出预测模型研究[J].中国煤炭,2012,38(11):102-106.Qu F,An W C,Li Y Y,et al.Research on prediction model for coal and gas outburst based on GRA-SVM[J].CHINA COAL,2012,38(11):102-106.
[8]张克.基于地震演模拟和SVM的煤与瓦斯突出危险区预测研究[D].中国矿业大学,2011,40-76.Zhang K.Research on Prediction of Coal and Gas Outburst Risk Area Based on Seismic Forward Modeling and SVM[D].China Mining University,2011,40-76.
[9]黄为勇,邵晓根,陈奎.一种采用CCPSO-SVM的煤与瓦斯突出预测方法[J].计算机科学,2012,39(11):216-225.Huang W Y,Shao X G,Chen K.Coal-and-Gas Outburst Forecast Using CCPSO and SVM[J].Computer Science,2012,39(11):216-225.
[10]刘俊娥,曾凡雷,郭章林.基于RS-SVM模型的煤与瓦斯突出多因素风险评价[J].中国安全科学学报,2011,21(7):21-25.Liu J E,Zeng F L,Guo Z L.Study on Multiple Factors Risk Evaluation of Coal and Gas Outburst Based on RS-SVM Model[J].China Safety Science Journal,2011,21(7):21-25.
[11]李润求,施式亮,念其锋.基于PSO-SVM煤矿瓦斯爆炸灾害风险模式识别[J].中国安全科学学报,2013,23(5):38-43.Li R Q,Shi S L,Nian Q F.Research on Pattern Recognition of Gas Explosion Disaster Risk in Coal Mines Based on PSO-SVM[J].China Safety Science Journal,2013,23(5):38-43.
[12]程东全,顾锋.基于可拓展模式识别煤与瓦斯突出危险性分析[J].安全与环境学报,2012,12(1):241-244.Chen D Q,Gu F.Risk analysis of coal and gas outburst based on extensible pattern recognition[J].Journal of Safety and Environment,2012,12(1):241-244.
[13]Hojjat Akhondi-Asl,James D.B.Nelson.M-Estimate robust PCAfor Seismic Noise Attenuation[C].IEEE International Conference on Image Processing,2016,1853-1857.
[14]Pan W T.A new fruit fly optimization algorithm:taking thefinancial distress model as an example[J].Knowledge-Based-Systems,2012,26:69-74.
[15]王雨虹,付华,张洋.基于KPCA和CIPSO-PNN的煤与瓦斯突出强度辨识模型[J].传感技术学报,2014,27(11):271-277.Wang Y H,Fu H,Zhang Y.The Identification Model of Coal and Gas OutburstIntensity Based on KPCA and CIPSO-PNN[J].CHINESE JOURNAL OF SENSORS AND ACTUATORS,2014,27(11):271-277.