摘要
台风灾情评估对防台减灾和救灾工作十分重要,为此,提出并建立了一种基于离散型Hopfield神经网络的评估模型。该模型选取5个评估因子,根据广东省台风灾害历史数据,按灾害等级划分原则制定了评估因子5个等级的划分方法,然后将待分级的台风进行Hopfield编码,建立离散型Hopfield神经网络模型,并用实例对模型进行了验证,分析模型的不足,将评估因子的权重因素引入到模型中,使模型更加合理。
The impact evaluation of typhoon disaster is very important to the work of typhoon emergence disaster re- duction and relief. This paper provides and makes a kind of evaluation model based on discrete Hopfield neural net- work. In this model we choose five evaluation factors. On the basis of typhoon disaster historical data for Guangdong Province and on the principle of disaster classification, we draw up the classification method of five grades of evalua- tion factors, then the typhoons were coded by Hopfield, and we make the discrete Hopfield neural network model and check the model with example. Finally, we analyze the shortage of the model, and then take the weight of the evaluation,factors of the model into consideration in order to make the model more rational.
出处
《自然灾害学报》
CSCD
北大核心
2011年第5期47-52,共6页
Journal of Natural Disasters
基金
广东省自然科学基金团队项目<华南沿海台风遥感检测与灾害评估>(8351030101000002)
广东省科技计划重点项目<重大台风灾害及城市火灾应急响应集成系统研制>(2010B031900041)
