有效预测震灾人员的存活情况是紧急配置应急资源和提高救援效率的首要工作。为提高震灾人员存活预测的精度，本文首先依据区域灾害系统理论和现有研究成果提出震灾人员存活预测指标。其次，针对震灾人员存活量指标数据的小样本、高维度、非线性特征，考虑将支持向量机（Support Vector Machine，SVM）模型引入震灾人员存活量预测中，为有效降低SVM在高维空间中非线性分类的误差，采用Mexican母小波核函数替换满足Mercer内积条件的核函数，以改变常规核函数缩小偏差的局限性，提出用于预测震灾人员存活量的Mexican小波SVM（Mexican Wavelet-SVM， Mexican Wv-SVM）模型。数值算例表明：相比于标准SVM、BP神经网络，Mexican Wv-SVM模型具有预测精度好、训练速度快和运行稳定性好的特征，证明了模型的可靠和有效。
The first work of distribution relief resource and improving the rescue efficiency is the survival amount prediction. The object of this paper is mainly to improve the prediction accuracy of the survival amount in earthquake disaster. First of all, the prediction indexes are put forward based on regional disaster theory and literatures. Secondly, the method of Support Vector Machine (SVM) model is introduced as the survival amount prediction in earthquake disaster to solve the index data of the small sample, high dimension and nonlinear characteristics. In this paper the model of the survival amount in earthquake disaster is put forward which replaced Mercer kernel function of inner product conditions to the Mexican mother Wavelet kernel function to effectively reduce the SVM classification of nonlinear error in high dimensional space and the limitations of conventional kernel function reducing the deviation. Finally, 53 groups of sample data are collected with the model test and these data came from the Chinese earthquake cases in 1989-2005. These sample data has the characteristics of small sample, nonlinear and high dimension that can be used to test the Mexican Wv-the SVM model. The numerical example shows Mexican Wv-the SVM model has high forecasting accuracy, fast training speed and running stability good characteristics to be compared with the standard SVM and BP neural network. In a word, the model is proved to be reliable and effective.
[1] Samardjieva E, Oike K.Modeling the number of casualties from earthquake[J].Journal of Natural Disaster Science, 1992,14( 1) : 17-28.
[2] Lomnitz C.Casualties and behavior of populati-ons during the earthquakes[J].Bulletin of the Seismologic Socilty of America, 1970,60(4):1309-1313.
[3] Tsai,et al.Spatial distribution and age dependence of human-fatality rates from the Chi-chi.Taiwan[J].Bulletin of the Seismological Societv of America,2001,91(5):1298-1309.
[4] 高惠英,李青霞.地震人员伤亡快速评估模型研究[J].灾害学,2010,25(1):275-277.
[5] 马玉宏,谢礼立.关于地震人员伤亡因素探讨[J].自然灾害学报.2000,9(3):84-90.
[6] 张洁,高惠英,刘琦.基于汶川地震的地震人员伤亡预测模型研究[J].中国安全科学学报,2011,21(3):59-64.
[7] 施伟华,陈坤华,谢英情,等.云南地震灾害人员伤亡预测方法研究[J].地震研究,2012,35(3): 387-392.
[8] 傅征祥,李革平. 地震生命损失研究[M].北京: 地震出版社,1993.
[9] Murakami H O. A simulation model to estimate human loss for occupants building in an earthquake [C]//Proeeding of the Tenth World Conference on Earthquake Engineering,Madrid,Spain,July 19-24,1992.
[10] 刘金龙,林均歧.基于震中烈度的地震人员伤亡评估方法研究[J].自然灾害学报,2012,21(5):113-119.
[11] 钱枫林.BP神经网络预测模型在应急物资需求中的应用——以地震伤亡人数预测为例[J].中国安全科学学报,2013,23(4):20-25.
[12] 何明哲,周文松.基于地震损伤指数的地震人员伤亡预测方法[J].哈尔滨工业大学学报,2011,43(4):23-27.
[13] 田鑫,朱冉冉.基于主成分分析及BP神经网络分析的地震人员伤亡预测模型[J].西北地震学报,2012,34(4):265-368.
[14] 吴恒景,冯铁男,洪中华,等.基于遥感图像的地震人员伤亡预测模型[J].同济大学学报(医学版),2013,34(5):36-39.
[15] Okada S, Indoor-zoning map on dwelling space safety during an earthquake[C]//Proceedings of the Tenth World Conference on Earthquake Engineering, Madrid.Spain,July 19-24,1992.
[16] 王斌,施祖见,匡蕾.基于改进KFDA和RWv-SVM的化工生产系统故障快速诊断[J].中国安全科学学报,2013,23(8):84-89.
[17] 张宏伟,朱自洁,宋卫华,等.基于改进的FOA-SVM导水裂隙带高度预测研究[J].中国安全科学学报,2013,23(10):9-14.
[18] 周建平,郑应平,王志萍.基于Morlet小波核多类支持向量机的故障诊断[J].华东电力, 2008, 36(8):952-955.
[19] 刘俊娥,安凤平,林大超,等.采煤工作面瓦斯涌出量的固有模态SVM建模预测[J].系统工程理论与实践,2013,33(2):505-511.
[20] 刁翠霞,陈思凤,刘业政.就与SVM求解不均衡数据集分类的主观权重约束方法[J].管理科学学报,2012,26(3):146-150.
[21] 张玲,陈涛,黄钧.基于最小最大后悔值的应急救援网络鲁棒优化模型与算法[J].中国管理科学,2014,22(7):131-139.
[22] Wu Qi, Yan Hongsen,Yang Hongbing. A hybrid forecasting model based on chaotic mapping and improved support vector machine[C]//Proceedings of the 9th International Conference for Yong Computer Scientists. Hunan,China,November 18-21,2008.
[23] Rossi F, Villab N. Support vector machine for functional data classification [J]. Neuron Computing, 2006, 69(7-9):730-742.
