基于粒子群优化最小二乘向量机的地震预测模型
|
摘要
为解决地震预测中最小二乘向量机(LSSVM)模型的参数难以确定的问题,利用粒子群算法(PSO)的收敛速度快和全局优化能力,优化LSSVM模型的惩罚因子和核函数参数,建立了PSO-LSSVM地震预测模型。通过对地震实例的预测仿真及其相关分析表明该方法的有效性。该方法优于传统的神经网络和支持向量机的地震预测方法,可以有效提高预测效能。
In order to overcome the problem of the uncertain parameters in LSSVM model,the PSO-LSSVM prediction model concerning earthquake forecast is developed,which is based on the particle swarm optimization algorithm with abilities of fast convergence and global optimization.The simulation results show that the proposed method is an effective tool for the prediction of earthquake,and it can effectively enhance the prediction accuracy compared with the way using neural network and support vector machine model.
In order to overcome the problem of the uncertain parameters in LSSVM model,the PSO-LSSVM prediction model concerning earthquake forecast is developed,which is based on the particle swarm optimization algorithm with abilities of fast convergence and global optimization.The simulation results show that the proposed method is an effective tool for the prediction of earthquake,and it can effectively enhance the prediction accuracy compared with the way using neural network and support vector machine model.
引文
[1]Ifantisk A,Giannakopoulos K.Changes of chaotic behavior ofthe long-term geoelectic potential difference observed during afive-year investigation and its possible relation to seismic activ-ity in Weatern Greece[J].Chaos,Solitons and Fractals,2002,14(5):779-795.
[2]Gotoh M,Hayakawa N A,Smirnova K.Fractal analysis ofseismogenic ULF emissions[J].Physics and Chemistry of theEarth,2004,29(4-9):419-424.
[3]King C,Azuma S,Ohno M,et al.In search of earthquakeprecursors in the water-level data of 16closely clustered wellsat Tono[J].International Journal Geophysical,2000,143(2):469-477.
[4]王炜,谢瑞,宋先月,等.使用人工神经网络进行我国大陆强震时间序列预测[J].西北地震学报,2002,24(4):315-320.
[5]陈超,曾三友,张好春,等.基于遗传神经网络的地震预测研究[J].计算机应用与软件,2008,25(4):135-137.
[6]陈以,王颖,张晋魁.组合人工神经网络在地震预测中的应用研究[J].计算机仿真,2011,28(1):190-193.
[7]Suykens J A K,Vandewalle J.Least squares support machineclassifier[J].Neural Processing Letters,1999,9(3):293-300.
[8]Van G T,Suykens J A K.Financial time series prediction u-sing least squares support vector machine within the evidenceframework[J].IEEE Transactions on Neural Networks,2001,12(4):809-821.
[9]Esen H,Ozgen F,Esen M.Modeling of a new solar air heaterthrough least squares support vector machine[J].Expert Sys-tems with Applications,2009,36(7):10673-10682.
[10]张春晓,张涛.基于最小二乘向量机和遗传算法的热式油水两相流含油率建模[J].化工学报,2009,60(7):1651-1655.
[11]龙文,梁昔明,董淑华,等.动态调整惯性权重的粒子群优化算法[J].计算机应用,2009,29(8):2240-2242.
[12]朱凯,王正林.精通MATLAB神经网络[M].北京:电子工业出版社,2010.
[13]王炜,林命,马钦忠,等.主成分分析法在地震预测中的应用研究[J].中国地震,2005,21(3):409-416.
[2]Gotoh M,Hayakawa N A,Smirnova K.Fractal analysis ofseismogenic ULF emissions[J].Physics and Chemistry of theEarth,2004,29(4-9):419-424.
[3]King C,Azuma S,Ohno M,et al.In search of earthquakeprecursors in the water-level data of 16closely clustered wellsat Tono[J].International Journal Geophysical,2000,143(2):469-477.
[4]王炜,谢瑞,宋先月,等.使用人工神经网络进行我国大陆强震时间序列预测[J].西北地震学报,2002,24(4):315-320.
[5]陈超,曾三友,张好春,等.基于遗传神经网络的地震预测研究[J].计算机应用与软件,2008,25(4):135-137.
[6]陈以,王颖,张晋魁.组合人工神经网络在地震预测中的应用研究[J].计算机仿真,2011,28(1):190-193.
[7]Suykens J A K,Vandewalle J.Least squares support machineclassifier[J].Neural Processing Letters,1999,9(3):293-300.
[8]Van G T,Suykens J A K.Financial time series prediction u-sing least squares support vector machine within the evidenceframework[J].IEEE Transactions on Neural Networks,2001,12(4):809-821.
[9]Esen H,Ozgen F,Esen M.Modeling of a new solar air heaterthrough least squares support vector machine[J].Expert Sys-tems with Applications,2009,36(7):10673-10682.
[10]张春晓,张涛.基于最小二乘向量机和遗传算法的热式油水两相流含油率建模[J].化工学报,2009,60(7):1651-1655.
[11]龙文,梁昔明,董淑华,等.动态调整惯性权重的粒子群优化算法[J].计算机应用,2009,29(8):2240-2242.
[12]朱凯,王正林.精通MATLAB神经网络[M].北京:电子工业出版社,2010.
[13]王炜,林命,马钦忠,等.主成分分析法在地震预测中的应用研究[J].中国地震,2005,21(3):409-416.
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