模糊神经网络预报模型的研究  

Forecast Model Study Based on Fuzzy Neural Network

作　　者：阳武[1] 伍元[1] 吴中如[1] 

机构地区：[1]河海大学水利水电工程学院

出　　处：《水电能源科学》2004年第1期63-65,共3页Water Resources and Power

基　　金：国家自然科学基金重点项目(50139030);教育部"跨世纪优秀人才培养计划"基金资助(2003512643)。

摘　　要：模糊系统与神经网络都能实现非线性的输入输出关系,但各有其优缺点。在对两者结合可行性和结合形式进行论述的基础上,探讨了采用结构等价型模糊神经网络建立大坝安全监控预报模型的方法和步骤,并将其应用于对某大坝变形的预测预报。实例分析表明,该模型预测结果优于常规统计模型,且模型的学习能力和表达能力均很强。Though either the fuzzy system or the neural network model can realize the nonlinear in/out relation, there are advantage and shortcoming themselves. By analyzing the feasibility and mode combining fuzzy system and neural network, the fuzzy neural network of equipotential structure is used to establish the forecast model of dam safety monitoring. The theory and method of this model and its application are studied in this paper. The example shows that the forecast precision of fuzzy neural network model is better than traditional statistic model, the learning and expressing ability of this combined model is strong.

关 键 词：大坝安全 神经网络 模糊系统 预报模型 

分 类 号：TV698.1[水利工程—水利水电工程]