地震波阻抗反演的免疫粒子群算法
|
摘要
提出了求解地震反演问题的免疫粒子群优化算法(IPSO).定义了种群多样性指标,在此基础上提出了抗体动态选择机制,分别在进化初期和后期采用不同策略进行抗体选择,与适应度选择、q-选择以及确定性选择方法比较优势明显.提出了基于种群多样性的权重自适应调整策略,在进化初期进行大范围全局搜索,进化后期则在相对较小的空间进行局部精细搜索,缓解了全局搜索和局部搜索之间的矛盾.二十层理论模型计算表明,IPSO算法不依赖于初始模型选择,在无噪和加入20%噪声情况下反演精度均明显优于标准粒子群算法(PSO)和自适应粒子群算法(APSO);成庄煤矿四、五盘区实际地震资料反演结果表明,IPSO算法能够准确识别煤层及顶底板位置,明显提高了弱反射波的连续性和可检测性.
Immune particle swarm optimization algorithm(IPSO) was proposed for the seismic inversion problem.The dynamic selection mechanism of antibodies based on the definition of swarms diversity was advanced.The mechanism is proved to be superior to the fitness selection,q-selection and confirmation selection by using different strategies in different evolution stages.The self-adjustment of the weight based on swarms diversity was put forward to relieve the contradiction between the global convergence and the local convergence by global search in the early stage and local search in the late stage.The simulation results indicate that the proposed IPSO has better efficiency and higher accuracy than PSO and APSO.The practical inversion results show that the inversion resolution is obviously higher than the seismic resolution,the continuity and the detection capability of weak reflection are improved greatly and the roof and floor of coal seam can be correctly identified.
Immune particle swarm optimization algorithm(IPSO) was proposed for the seismic inversion problem.The dynamic selection mechanism of antibodies based on the definition of swarms diversity was advanced.The mechanism is proved to be superior to the fitness selection,q-selection and confirmation selection by using different strategies in different evolution stages.The self-adjustment of the weight based on swarms diversity was put forward to relieve the contradiction between the global convergence and the local convergence by global search in the early stage and local search in the late stage.The simulation results indicate that the proposed IPSO has better efficiency and higher accuracy than PSO and APSO.The practical inversion results show that the inversion resolution is obviously higher than the seismic resolution,the continuity and the detection capability of weak reflection are improved greatly and the roof and floor of coal seam can be correctly identified.
引文
[1]杨文采.评地球物理反演的发展趋向[J].地学前缘,2002,9(4):389-396.YANG Wen-cai.Trends of geophysics inversion[J].Earth Science Frontiers,2002,9(4):389-396.
[2]姚姚.地球物理非线性反演模拟退火法的改进[J].地球物理学报,1995,38(5):643-650.YAO Yao.SAi mprovement of geophysics nonlinear inversion[J].Chinese Journal of Geophysics,1995,38(5):643-650.
[3]罗红明,王家映,朱培民.基于免疫算法的地球物理反演研究[J].石油地球物理勘探,2008,43(2):222-228.LUO Hong-ming,WANG Jia-ying,ZHU Pei-min.Study of geophysical inversion based on i mmunity al-gorithm[J].Oil Geophysical Prospecting,2008,43(2):222-228.
[4]蔡涵鹏,贺振华,黄德济.基于粒子群优化算法波阻抗反演研究与应用[J].石油地球物理勘探,2008,43(5):535-539.CAI Han-peng,HE Zhen-hua,HUANG De-ji.Wave i mpedance inversion based on PSO[J].Oil Geophysi-cal Prospecting,2008,43(5):535-539.
[5]袁三一,陈小宏.一种新的地震子波与层速度反演方法[J].地球物理学进展,2008,23(1):198-205.YUAN San-yi,CHEN Xiao-hong.Anew method for seismic wavelet extraction and interval velocity inver-sion[J].Progress in Geophysics,2008,23(1):198-205.
[6]ZHIGANG L,XINGSHENG G,BIN J.A novel particle swarm opti mization algorithm for permuta-tion flow-shop scheduling to mini mize makespan[J].Chaos,Solitons&Fractals,2008,35(5):851-861.
[7]COELHO LS,LEE C S.Solving economic load dis-patch problems in power systems using chaotic and Gaussian particle swarmopti mization approaches[J].Electrical Power&Energy Systems,2008,30(5):297-307.
[8]LIN Cheng-jian.An efficient i mmune-based symbiot-ic particle swarm opti mization learning algorithmfor TSK-type neuro-fuzzy networks design[J].Fuzzy Sets and Systems,2008,159(21):2890-2909.
[9]GE Hong-wei,SUN Liang,LIANG Yan-chun.An effective PSO and AIS-based hybrid intelligent algo-rithmfor job-shop scheduling[J].IEEE Transactions on Systems,Man and Cybernetics-Part A:Systems and Humans,2008,38(2):358-368.
[2]姚姚.地球物理非线性反演模拟退火法的改进[J].地球物理学报,1995,38(5):643-650.YAO Yao.SAi mprovement of geophysics nonlinear inversion[J].Chinese Journal of Geophysics,1995,38(5):643-650.
[3]罗红明,王家映,朱培民.基于免疫算法的地球物理反演研究[J].石油地球物理勘探,2008,43(2):222-228.LUO Hong-ming,WANG Jia-ying,ZHU Pei-min.Study of geophysical inversion based on i mmunity al-gorithm[J].Oil Geophysical Prospecting,2008,43(2):222-228.
[4]蔡涵鹏,贺振华,黄德济.基于粒子群优化算法波阻抗反演研究与应用[J].石油地球物理勘探,2008,43(5):535-539.CAI Han-peng,HE Zhen-hua,HUANG De-ji.Wave i mpedance inversion based on PSO[J].Oil Geophysi-cal Prospecting,2008,43(5):535-539.
[5]袁三一,陈小宏.一种新的地震子波与层速度反演方法[J].地球物理学进展,2008,23(1):198-205.YUAN San-yi,CHEN Xiao-hong.Anew method for seismic wavelet extraction and interval velocity inver-sion[J].Progress in Geophysics,2008,23(1):198-205.
[6]ZHIGANG L,XINGSHENG G,BIN J.A novel particle swarm opti mization algorithm for permuta-tion flow-shop scheduling to mini mize makespan[J].Chaos,Solitons&Fractals,2008,35(5):851-861.
[7]COELHO LS,LEE C S.Solving economic load dis-patch problems in power systems using chaotic and Gaussian particle swarmopti mization approaches[J].Electrical Power&Energy Systems,2008,30(5):297-307.
[8]LIN Cheng-jian.An efficient i mmune-based symbiot-ic particle swarm opti mization learning algorithmfor TSK-type neuro-fuzzy networks design[J].Fuzzy Sets and Systems,2008,159(21):2890-2909.
[9]GE Hong-wei,SUN Liang,LIANG Yan-chun.An effective PSO and AIS-based hybrid intelligent algo-rithmfor job-shop scheduling[J].IEEE Transactions on Systems,Man and Cybernetics-Part A:Systems and Humans,2008,38(2):358-368.
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