基于神经网络的高炉异常炉况判断专家系统

Neuron-Based Expert System for Judging the State of Blast Furnace

摘要: 采用反向传播网络作为推理机,构造了高炉异常炉况判断专家系统,该系统具有良好的自学习功能和联想记忆功能.系统采用离线学习方式,在线运行时,可将高炉操作实绩存入知识库,作为进一步学习的样本,从而提高了系统精度和联想能力.

Abstract: An expert system based on neural network was establised for judging the state of blast furnace.The Back-Propagation network is used as engine of ES.This ES with selflearning function and associative memory function can be learning when the system off line and can save the samples into the knowledge base when the system on line.So the system ability can be developed continuously.