神经网络系统理论在换热器分配特性研究中的应用

The Neural Network Application in Maldistribution Analysis of Plate Fin Heat Exchanger.

本文采用神经网络系统理论ＢＰ算法，实现换热器内部流量分配状态的评估，即均匀、微弱不均匀、不均匀、严重不均匀四种状态的学习和分类识别。该算法输出收效，学习时间为２００秒时，错误识别率在８％以内，可用于在线监控。

A artificial nearal network is proposed for the flow distribution analysis of plate fin heat exchanger.In the neural network,those system variables which play an important role in distributiion in exchanger such as inlet pressure P and flow rate Q are employed as the inputs.The output of the neural network provides the information on that the distribution is good,fine,bad or worse.Once the connection weights of the neural network have been learned using a set of tranining data,the neural net can be applied to analyze the flow conditin in real time where the operating condition changes with time.It can be seen as a wholly new method which has bright prospects.