摘要
文章提出了一种基于模糊规则的神经网络结构,并用形式化语言进行描述。基于模糊规则的神经网络由输入层、规则层和输出层三层网络结构组成,以隶属度函数(语义值)作为网络权值,输入值沿权值的传播即进行隶属度计算。在充分分析三角形函数特征的基础上,应用启发式方法,导出了FRBNN网络的学习算法。最后应用FRBNN评价船舶碰撞危险度,表明FRBNN兼备神经网络和模糊推理系统的优点。
: An architecture for FRBNN(Fuzzy Rule-based Neural Network)is proposed,and applying a descriptive language for expressing it.FRBNN is composed of input layer,rule layer and output layer,and use membership function as net connection weights.The propagation of input value along the fuzzy weight simply result in membership value.After analyzing the characteristic of triangular function,This paper derives heuristic learning algorithm for FRBNN.Then It applies FRBNN for collision risk accessment,the result shows that FRBNN has the advantage of neural network and fuzzy inference systems.
出处
《计算机工程与应用》
CSCD
北大核心
2001年第5期6-8,共3页
Computer Engineering and Applications
基金
国家自然科学基金资助