【摘要】 文章尝试将模糊神经网络方法引入ATM网络的业务量预测中。ATM网络业务源一般是随机产生的时变信号，其模型一般很难描述。文章充分考虑了模糊神经网络的学习功能，通过对相关模型的仿真，能够很好地描述ATM网络中的业务流特性，对多媒体的业务量做出了准确的预测。与传统的神经网络方法比较，具有更好的逼近效果。更多还原

【Abstract】 This paper brings the fuzzy neural network algorithms into the traffic prediction in ATM networks．In general，ATM traffic source is a stochastic time variant signal．The model is difficult to describe．This paper concerns about the learning function of the fuzzy neural network，by simulation It can describe and predict the source characteristic in ATM networks with accuracy．As a result，This fuzzy neural network algorithm has better effect．更多还原