摘要
预测鱼雷罐车(TPC)在高炉的实际受铁量,对协调铁水平衡、减少兑罐次数和温降损失,保证高炉出铁安全,提高TPC利用率具有重要作用。采用主成分分析(PCA)提取过程特征参数,并剔除相关冗余信息;BP神经网络用来逼近受铁量预测这一非线性过程;改进了遗传算法(GA)的适应度函数,并精确给定BP神经网络的权值和阈值,进而建立了基于PCA-GA-BP的TPC受铁量预测模型。采用某钢铁企业实际生产数据运算,结果表明模型合理、有效,提高了鱼雷罐车(TPC)受铁量预测准确性。
Prediction problem of torpedo ladle car ( TPC ) actual reception iron amount under the blast furnace is discussed, which has an important function to coordinate molten iron balance, decrease exchanging bottles, reduce temperature loss and raise TPC utilization. A prediction model of TPC reception iron amount based on PCA-GA-BP is proposed. The principle component analysis is used to seleet the most relevant process features and to eliminate the con'elations of the input variables. Back-propagation neural network is used to characterize the nonlinearity and accuracy. Genetic algorithm is employed to optimize the parameters and structure of the BP neural net- work by improving GA fitness function. Experiment results through the actual production data of an enterprise show the prediction rationality and validity, and the prediction accuracy of TPC actual reception iron amount is increased.
出处
《控制工程》
CSCD
北大核心
2009年第4期446-450,共5页
Control Engineering of China
基金
国家973重大基础研究计划基金资助项目(2002CB312201)
国家863高技术研究计划基金资助项目(2004AA412030)
关键词
受铁量
鱼雷罐车
主成分分析
遗传算法
BP神经网络
reception iron amount
torpedo ladle car
principle component analysis
genetic algorithm
back-propagation neural network