基于离散Hopfield神经网络的突发事件连锁反应路径推演模型
|
摘要
突发事件及其引发的连锁反应会造成严重的灾害后果,如果事先知道突发事件可能引发的连锁反应路径,就可以提前做好断链减灾工作.因此,提出一种基于Hopfield神经网络的突发事件连锁反应路径推演模型.该模型用Hopfield神经网络表示一般的突发事件网络,用Hopfield神经网络的运行规则表示突发事件连锁反应的原理,并设置神经元阈值,将突发事件连锁反应路径的推演过程映射为Hopfield神经网络的演化过程,通过运行Hopfield神经网络推演初始突发事件的一条连锁反应路径.最后,用该模型推演了地震引发的连锁反应路径,验证了该模型的合理性.
Emergency events and the resulting chain reactions would cause serious consequences,i f the chain reaction path was known in advance, the chain would be interrupted and the disaster reduced.To solve this problem, an emergency event chain reaction path deduction model based on discrete Hopfield neural network is presented.The model uses the Hopfield neural network to represent a normal emergency event network, then deduct the chain reaction path of the emergency event according to the principle of the chain reaction, and by setting the threshold of the neurons,i t maps the inference process of emergency event chain reaction to the evolution process of Hopfield neural network, and the chain reaction path of initial emergency event can be found out by running the Hopfield neural network.Finally, the rationality of this method is validated by analyzing the resulting chain reactions of earthquake.
Emergency events and the resulting chain reactions would cause serious consequences,i f the chain reaction path was known in advance, the chain would be interrupted and the disaster reduced.To solve this problem, an emergency event chain reaction path deduction model based on discrete Hopfield neural network is presented.The model uses the Hopfield neural network to represent a normal emergency event network, then deduct the chain reaction path of the emergency event according to the principle of the chain reaction, and by setting the threshold of the neurons,i t maps the inference process of emergency event chain reaction to the evolution process of Hopfield neural network, and the chain reaction path of initial emergency event can be found out by running the Hopfield neural network.Finally, the rationality of this method is validated by analyzing the resulting chain reactions of earthquake.
引文
[1]裘江南,王延章,董磊磊,等.基于贝叶斯网络的突发事件预测模型[J].系统管理学报,2011,20(1):98-103.QIU Jiang-nan,WANG Yan-zhang,DONG Lei-lei,et al.A model for predicting emergency event basedon Bayesian networks[J].Journal of Systems&Management,2011,20(1):98-103.(in Chinese)
[2]方志耕,杨保华,陆志鹏,等.基于Bayes推理的灾害演化GERT网络模型研究[J].中国管理科学,2009,17(2):102-107.FANG Zhi-geng,YANG Bao-hua,LU Zhi-peng,et al.The GERT network model study of disasterevolution based on Bayes inference[J].ChineseJournal of Management Science,2009,17(2):102-107.(in Chinese)
[3]Rousseau G,Giorgini B,Livi R.Dynamical phasesin a cellular automation model for epidemicpropagation[J].Physica D:Nonlinear Phenomena,1997,103(1):554-563.
[4]闪淳昌.建立健全突发事件应急预案[J].中国减灾,2004(6):15-16.SHAN Chun-chang.The establishment of a soundemergency plan[J].Disaster Reduction in China,2004(6):15-16.(in Chinese)
[5]王建伟,荣莉莉.突发事件的连锁反应网络模型研究[J].计算机应用研究,2008,25(11):3288-3291.WANG Jian-wei,RONG Li-li.Research on chain-reacting network model of emergency events[J].Application Research of Computers,2008,25(11):3288-3291.(in Chinese)
[6]荣莉莉,蔡莹莹,王铎.基于共现分析的我国突发事件关联研究[J].系统工程,2011,29(6):1-7.RONG Li-li,CAI Ying-ying,WANG Duo.Relations of emergency events occurring in Chinabased on the co-occurrence analysis[J].SystemsEngineering,2011,29(6):1-7.(in Chinese)
[7]彭岩,王万森,王旭仁,等.基于机器学习的风险预测方法研究[J].计算机科学,2009,36(4):205-210.PENG Yan,WANG Wan-sen,WANG Xu-ren,et al.Research of machine learning based riskprediction[J].Computer Science,2009,36(4):205-210.(in Chinese)
[8]Hopfield J J.Neural network and physical systemswith emergent collective computational abilities[J].Proceedings of the National Academy of Science,1982,79(4):2554-2558.
[9]Kishimoto M,Sakasai K,Ara K.Solution ofelectromagnetic inverse problem using combinationalmethod of Hopfield neural network and geneticalgorithm[J].Journal of Application Physics,1996,79(1):1-7.
[10]Kim Tae-hyung,Kim Yoo-shin.A study on thegeneralized model of statistical Hopfield neuralnetwork to solve the combinational optimizationproblem[J].Journal of the Institute of ElectronicsEngineers of Korea C,1999,36(10):66-74.
[11]盛秋艳,何文广.基于Hopfield神经网络的概念检索技术[J].情报科学,2004,22(2):346-348.SHENG Qiu-yan,HE Wen-guang.Research on theconcept retrieval technology based on the Hopfieldnervous network[J].Information Science,2004,22(2):346-348.(in Chinese)
[12]LIU De-rong,LU Zan-jun.Some properties forimproved information storage and retrieval inHopfield-type neural networks[C]//Proceedingsof the 36th IEEE Conference on Decision andControl.California:IEEE Control Systems Society,1997:486-491.
[13]Chen H.A machine learning approach to documentretrieval:An overview and an experiment[C]//Proceedings of the Twenty-Seventh Annual HawaiiInternational Conference on System Sciences.California:IEEE Computer Society Press,1994:631-640.
[14]Chen H,Ng T.An algorithmic approach to conceptexploration in a large knowledge network symbolicbranch-and-bound search VS connectionist Hopfieldnet activation[J].Journal of the American Societyfor Information Science,1995,46(5):348-369.
[15]Chuang Shang-jen,Chen Chiung-hsing,Kao Chien-chih.Improving pattern recognition rate byGaussian Hopfield neural network[J].AdvancedMaterials Research,2011,189-193:2042-2045.
[16]王铎.基于关联度的突发事件网络模型研究[D].大连:大连理工大学,2010.WANG Duo.Network model of emergency eventsbased on correlation[D].Dalian:Dalian Universityof Technology,2010.(in Chinese)
[17]陈运泰,杨智娴,许力生.海啸、地震海啸与海啸地震[J].物理,2005,34(12):864-872.CHEN Yun-tai,YANG Zhi-xian,XU Li-sheng.Tsunamis,earthquake-generated tsunamis andtsunamigenic earthquakes[J].Physics,2005,34(12):864-872.(in Chinese)
[18]龚茂珣,堵盘军,薛志刚.上海沿海海洋灾害的危害及应对措施:缅甸特大风暴潮灾害的启示[J].华东师范大学学报:自然科学版,2008(5):126-133.GONG Mao-xun,DU Pan-jun,XUE Zhi-gang.Marine hazards in Shanghai coastal regions andcorresponding countermeasures:Lessons from thestorm surge disaster of Burma[J].Journal of EastChina Normal University:Natural Science,2008(5):126-133.(in Chinese)
[19]李鑫,章卫胜,张金善,等.一次典型寒潮风暴潮过程的数值模拟研究[J].海洋科学进展,2010,28(1):8-16.LI Xin,ZHANG Wei-sheng,ZHANG Jin-shan,et al.Study on a typical storm surge process causedby cold air[J].Advances in Marine Science,2010,28(1):8-16.(in Chinese)
[2]方志耕,杨保华,陆志鹏,等.基于Bayes推理的灾害演化GERT网络模型研究[J].中国管理科学,2009,17(2):102-107.FANG Zhi-geng,YANG Bao-hua,LU Zhi-peng,et al.The GERT network model study of disasterevolution based on Bayes inference[J].ChineseJournal of Management Science,2009,17(2):102-107.(in Chinese)
[3]Rousseau G,Giorgini B,Livi R.Dynamical phasesin a cellular automation model for epidemicpropagation[J].Physica D:Nonlinear Phenomena,1997,103(1):554-563.
[4]闪淳昌.建立健全突发事件应急预案[J].中国减灾,2004(6):15-16.SHAN Chun-chang.The establishment of a soundemergency plan[J].Disaster Reduction in China,2004(6):15-16.(in Chinese)
[5]王建伟,荣莉莉.突发事件的连锁反应网络模型研究[J].计算机应用研究,2008,25(11):3288-3291.WANG Jian-wei,RONG Li-li.Research on chain-reacting network model of emergency events[J].Application Research of Computers,2008,25(11):3288-3291.(in Chinese)
[6]荣莉莉,蔡莹莹,王铎.基于共现分析的我国突发事件关联研究[J].系统工程,2011,29(6):1-7.RONG Li-li,CAI Ying-ying,WANG Duo.Relations of emergency events occurring in Chinabased on the co-occurrence analysis[J].SystemsEngineering,2011,29(6):1-7.(in Chinese)
[7]彭岩,王万森,王旭仁,等.基于机器学习的风险预测方法研究[J].计算机科学,2009,36(4):205-210.PENG Yan,WANG Wan-sen,WANG Xu-ren,et al.Research of machine learning based riskprediction[J].Computer Science,2009,36(4):205-210.(in Chinese)
[8]Hopfield J J.Neural network and physical systemswith emergent collective computational abilities[J].Proceedings of the National Academy of Science,1982,79(4):2554-2558.
[9]Kishimoto M,Sakasai K,Ara K.Solution ofelectromagnetic inverse problem using combinationalmethod of Hopfield neural network and geneticalgorithm[J].Journal of Application Physics,1996,79(1):1-7.
[10]Kim Tae-hyung,Kim Yoo-shin.A study on thegeneralized model of statistical Hopfield neuralnetwork to solve the combinational optimizationproblem[J].Journal of the Institute of ElectronicsEngineers of Korea C,1999,36(10):66-74.
[11]盛秋艳,何文广.基于Hopfield神经网络的概念检索技术[J].情报科学,2004,22(2):346-348.SHENG Qiu-yan,HE Wen-guang.Research on theconcept retrieval technology based on the Hopfieldnervous network[J].Information Science,2004,22(2):346-348.(in Chinese)
[12]LIU De-rong,LU Zan-jun.Some properties forimproved information storage and retrieval inHopfield-type neural networks[C]//Proceedingsof the 36th IEEE Conference on Decision andControl.California:IEEE Control Systems Society,1997:486-491.
[13]Chen H.A machine learning approach to documentretrieval:An overview and an experiment[C]//Proceedings of the Twenty-Seventh Annual HawaiiInternational Conference on System Sciences.California:IEEE Computer Society Press,1994:631-640.
[14]Chen H,Ng T.An algorithmic approach to conceptexploration in a large knowledge network symbolicbranch-and-bound search VS connectionist Hopfieldnet activation[J].Journal of the American Societyfor Information Science,1995,46(5):348-369.
[15]Chuang Shang-jen,Chen Chiung-hsing,Kao Chien-chih.Improving pattern recognition rate byGaussian Hopfield neural network[J].AdvancedMaterials Research,2011,189-193:2042-2045.
[16]王铎.基于关联度的突发事件网络模型研究[D].大连:大连理工大学,2010.WANG Duo.Network model of emergency eventsbased on correlation[D].Dalian:Dalian Universityof Technology,2010.(in Chinese)
[17]陈运泰,杨智娴,许力生.海啸、地震海啸与海啸地震[J].物理,2005,34(12):864-872.CHEN Yun-tai,YANG Zhi-xian,XU Li-sheng.Tsunamis,earthquake-generated tsunamis andtsunamigenic earthquakes[J].Physics,2005,34(12):864-872.(in Chinese)
[18]龚茂珣,堵盘军,薛志刚.上海沿海海洋灾害的危害及应对措施:缅甸特大风暴潮灾害的启示[J].华东师范大学学报:自然科学版,2008(5):126-133.GONG Mao-xun,DU Pan-jun,XUE Zhi-gang.Marine hazards in Shanghai coastal regions andcorresponding countermeasures:Lessons from thestorm surge disaster of Burma[J].Journal of EastChina Normal University:Natural Science,2008(5):126-133.(in Chinese)
[19]李鑫,章卫胜,张金善,等.一次典型寒潮风暴潮过程的数值模拟研究[J].海洋科学进展,2010,28(1):8-16.LI Xin,ZHANG Wei-sheng,ZHANG Jin-shan,et al.Study on a typical storm surge process causedby cold air[J].Advances in Marine Science,2010,28(1):8-16.(in Chinese)
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