摘要
根据管式裂解炉的特点， 提出了一种管式裂解炉出口温度优化控制方法。给出了作为模型预估器的神经网络ＧＡ—ＢＰ算法流程及ＧＡ 算法实现， 提出了最优控制指标选择原则及控制指标表达式。经计算机对四组裂解管生产乙烯的工业对象进行仿真研究表明， 该控制方法具有良好的跟踪性能及抗干扰能力。
According to the characteristic of the pipeline cracking furnace, a new neural network optimization control strategy of cracking furnace′s output temperature is put forward. The GA-BP learning algorithm of neural network, the GA learning algorithm, the rule of optimum control including their features were introduced. The computer simulation has proved that the anti-jamming capacity and the track performance of the strategy is good.
出处
《江苏石油化工学院学报》
1999年第3期40-43,共4页
Journal of Jiangsu Institute of Petrochemical Technology
关键词
管式
裂解炉
神经网络
温度控制
炼油
Pipeline cracking furnace
Neural network
Temperature control
