基于PSO-SVM的煤岩声发射源定位预测
|
摘要
为了准确的预测采空区煤矿煤岩破裂与失稳前,岩石所释放出来的声发射信息的位置,并且根据山西焦煤的官地矿16403工作面获得的声发射事件的数据,因为该数据是一个非线性、高维的问题,提出了用PSO和SVM算法相结合的方法应用在煤岩声发射源定位中。以往的方法只是单纯的收集煤岩或岩石声发射信息,以至于定位出现失准、精度低和误差大的缺点。基于煤岩失稳前两者都会发出强烈的信号,文章提出了"1+1=1"的定位模型,既收集同一位置的岩石和煤岩体的声发射信号,由PSO-SVM分析处理后,得到其位置。仿真结果表明:应用PSO和SVM结合的方法进行煤岩声发射源定位的预测,在提高准确性和精确度的同时,也大大的提高了泛化的能力,该方法也大大减小定位失准的误差。
In order to predict accurately acoustic emission information location released by rock before coal rock fracture and instability,and the acoustic emission event data is a non-linear and high-dimensional problem gained from 16403 working zone in Guandi coal Shanxi Coking,The method of combining PSO and SVM algorithm used in the positioning of coal and rock acoustic emission in this paper.The traditional method has a disadvantage of misalignment,low accuracy and big error,because it simply collect coal rock or rock acoustic emission information.According to strong signal sent by the two,the paper proposes a"1+1=1"positioning method,in other words,the location is confirmed by collecting the rock and coal-rock acoustic emission information processed by PSO-SVM in the same location.The simulation results show that this method improves accuracy and precision,at the same time,the generalization ability is improved greatly,it can also greatly reduce inaccurate positioning error.
In order to predict accurately acoustic emission information location released by rock before coal rock fracture and instability,and the acoustic emission event data is a non-linear and high-dimensional problem gained from 16403 working zone in Guandi coal Shanxi Coking,The method of combining PSO and SVM algorithm used in the positioning of coal and rock acoustic emission in this paper.The traditional method has a disadvantage of misalignment,low accuracy and big error,because it simply collect coal rock or rock acoustic emission information.According to strong signal sent by the two,the paper proposes a"1+1=1"positioning method,in other words,the location is confirmed by collecting the rock and coal-rock acoustic emission information processed by PSO-SVM in the same location.The simulation results show that this method improves accuracy and precision,at the same time,the generalization ability is improved greatly,it can also greatly reduce inaccurate positioning error.
引文
[1]Lockner D.The Role of Acoustic Emission in the Study of RockFracture[J].International Journal of Rock Mechanics and MiningSciences and Geomechanics Abstracts,1993,30(7):883-899.
[2]李世愚,和雪松,张天中,等.地震学在减轻矿山地质灾害中的应用进展[J].国际地震动态,2006(4):2-9.
[3]成新民,胡峰,邓艾东,等.基于代数网络和混沌参数的碰摩声发射源定位方法研究[J].振动工程学报,2011,24(3):287-293.
[4]康玉梅,刘建坡,李海滨,等.一类基于最小二乘法的声发射源组合定位算法[J].东北大学学报,2010,31(11):1648-1656.
[5]Geiger L.Probability Method for the Determination of EarthquakeEpicenters from the Arrival Time Only[J].Bull StLouis Unic,1912,8:60-71.
[6]Geiger L.Herbsetimmung Bei Erdbeben Aus Den Ankunfzeiten[J].Gessell Wiss Goett,1910,4:331-349.
[7]樊保圣,闫小青,余丽,等.基于Lamb波频散特性的薄板声发射源定位方法研究[J].固体力学学报,2011,32:283-287.
[8]杨永杰,王德超,陈绍杰,等.基于离散小波分析的灰岩压缩破坏声发射预测研究[J].煤炭学报,2010,35(2):213-217.
[9]李怀珍,袁瑞甫.基于神经网络的声发射定位技术及在岩石破裂实验中的应用[J].矿业工程,2005,3(5):3-5.
[10]Rosenstern M T,Collins J J,Deluca C J.A Practical Method forCalculating Largest Lyapunov Exponents from Small Data Sets[J].Physical D,1993,65:117-134.
[11]王妍,徐伟,曲继圣.基于时间序列的相空间重构算法及验证(一)[J].山东大学学报,2005,35(4):109-114.
[12]Grassberger P,Procaccia L.Measuring the Strangeness if StrangeAttractors[J].Physica D,1983,9:189-208.
[13]张朝龙,江巨浪,李彦梅,等.基于云粒子群-最小二乘支持向量机的传感器温度补偿[J].传感技术学报,2012,25(4):472-477.
[14]王涛,王洋洋,郭长娜,等.QGA-RBF神经网络在矿井瓦斯涌出量预测中的应用[J].传感技术学报,2012,25(1):119-123.
[15]黄存捍,冯涛,王卫军,等.基于分形和支持向量机矿井涌水量的预测[J].煤炭学报,2010,35(5):806-810.
[2]李世愚,和雪松,张天中,等.地震学在减轻矿山地质灾害中的应用进展[J].国际地震动态,2006(4):2-9.
[3]成新民,胡峰,邓艾东,等.基于代数网络和混沌参数的碰摩声发射源定位方法研究[J].振动工程学报,2011,24(3):287-293.
[4]康玉梅,刘建坡,李海滨,等.一类基于最小二乘法的声发射源组合定位算法[J].东北大学学报,2010,31(11):1648-1656.
[5]Geiger L.Probability Method for the Determination of EarthquakeEpicenters from the Arrival Time Only[J].Bull StLouis Unic,1912,8:60-71.
[6]Geiger L.Herbsetimmung Bei Erdbeben Aus Den Ankunfzeiten[J].Gessell Wiss Goett,1910,4:331-349.
[7]樊保圣,闫小青,余丽,等.基于Lamb波频散特性的薄板声发射源定位方法研究[J].固体力学学报,2011,32:283-287.
[8]杨永杰,王德超,陈绍杰,等.基于离散小波分析的灰岩压缩破坏声发射预测研究[J].煤炭学报,2010,35(2):213-217.
[9]李怀珍,袁瑞甫.基于神经网络的声发射定位技术及在岩石破裂实验中的应用[J].矿业工程,2005,3(5):3-5.
[10]Rosenstern M T,Collins J J,Deluca C J.A Practical Method forCalculating Largest Lyapunov Exponents from Small Data Sets[J].Physical D,1993,65:117-134.
[11]王妍,徐伟,曲继圣.基于时间序列的相空间重构算法及验证(一)[J].山东大学学报,2005,35(4):109-114.
[12]Grassberger P,Procaccia L.Measuring the Strangeness if StrangeAttractors[J].Physica D,1983,9:189-208.
[13]张朝龙,江巨浪,李彦梅,等.基于云粒子群-最小二乘支持向量机的传感器温度补偿[J].传感技术学报,2012,25(4):472-477.
[14]王涛,王洋洋,郭长娜,等.QGA-RBF神经网络在矿井瓦斯涌出量预测中的应用[J].传感技术学报,2012,25(1):119-123.
[15]黄存捍,冯涛,王卫军,等.基于分形和支持向量机矿井涌水量的预测[J].煤炭学报,2010,35(5):806-810.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心