感知器算法在台风风暴潮灾情等级评估中的应用  

The Application of the Perception Approach in the Evaluation of Typhoon Storm Surge Disaster Degree

作　　者：叶雯[1] 刘美南[1] 陈晓宏[1] 

机构地区：[1]中山大学水资源与环境系,广东广州510275

出　　处：《中山大学学报（自然科学版）》2004年第2期117-120,共4页Acta Scientiarum Naturalium Universitatis Sunyatseni

基　　金：国家自然科学基金资助项目(49971019)

摘　　要：风暴潮灾害每年都给我国沿海地区造成巨大的经济损失。在前人研究的基础上,引入风暴潮灾情等级的模式识别思想,运用感知器方法的理论及其算法,并以广东省湛江地区为例,建立了该地区的台风风暴潮灾情等级判别函数,经还原检验和判别预测获得满意的结果,证明该方法用于台风风暴潮灾情等级判别是可行的。这对于实现风暴潮灾情等级的快速评估,为抢险救灾的人力、财力和物力分配提供依据。Storm surge disasters result in great economic loss in coastal areas of China every year. The idea of pattern recognition is introduced into storm surge disaster degree evaluation for the first time. Using the theory and method of the perceptron approach and the loss data of Zhanjiang District of Guangdong Province,a discriminant function of typhoon storm surge disaster degree is established with satisfactory results. It is proved that the perceptron approach is feasible in storm surge disaster degree evaluation. It is significant in realizing rapid evaluation of storm surge disaster degree and distributing the manpower and material resources reasonably when providing disaster relief.

关 键 词：模式识别 台风风暴潮 灾情等级 

分 类 号：P426.62[天文地球—大气科学及气象学]