摘要
常压塔四线350℃馏出含量是炼油厂常压蒸馏生产过程的重要质量指标,它与常压炉出口温度等多个变量之间存在严重的非线性关系,而且无法实时在线用仪表直接测量。论文提出了基于RBF神经网络的常四350℃含量预报模型,并用计算机软测量方式,对中石化广州分公司常压蒸馏装置(一)的实际数据进行模型验证研究。实验结果表明,该方法速度快,对实际生产具有指导意义。
The 350℃ fraction of petrochemical distillation device is important to the quality of products.However,it has serious non-linearity affected by several parameters such as normal pressure furnace＇s exit temperature.And it is impossible to measure it by any online instruments.The prediction of petrochemical distillation device 350℃ fraction based on RBF neural network is put forward,and the research of verifying the model is made by comparing the predictive value with the practical data of Guangzhou Petrochemical Complex. The result shows this method is fast and makes sense to practice.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第30期224-226,共3页
Computer Engineering and Applications
基金
广东省科技计划资助项目(编号:2003B50301)
