摘要
洪水灾情评估实际上是一种模式识别问题，而人工神经网络模型具有逼近有界闭子集上任意非线性映射的特性，且模型更新方便．本文提出了基于遗传算法的洪水灾情评估神经网络模型，阐述了其基本原理和其法，实例研究表明其实用性、客观性和通用性．
Evaluation of flood disaster is a pattern recognition problem in fact. Theartificial neural network model has the characteristics of expressing arbitrary nonlinearmapping and can be renewed easily. In this paper, the improved genetic algorithm basedon neural network model (GANN) for evaluation of fIood disaster is worked out and theprinciple and algorithms are described. The application of GANN to evaluation of flooddisaster shows its practicality, objectivity and generality.
出处
《灾害学》
CSCD
1998年第2期6-11,共6页
Journal of Catastrophology
基金
国家"九五"科技攻关课题
关键词
洪水
灾情评估
神经网络
遗传算法
神经网络模型
Evaluation of flood disaster, Artificial neural network, Genetic algorithm
