论文部分内容阅读
随着乙烯裂解原料日趋多样化,液化石油气(LPG)的使用比重逐渐增大,而对于以LPG为原料的裂解过程建模与优化研究却较少,因此本文对以LPG为裂解原料的裂解过程进行建模与优化研究。利用裂解机理模拟LPG裂解过程,该过程同时考虑裂解原料的组分变化和操作条件变化对关键裂解产品收率的影响,然后根据实验设计原理产生适量仿真数据,建立了以原料组成与操作条件为输入、关键裂解产品收率为输出的PSOBP神经网络收率模型。通过模型验证,得到工业现场丙烯乙烯比值与本文模型预测得到的丙烯乙烯比的平均相对误差仅为2.142%。基于该模型利用PSO算法对LPG原料组成与操作条件进行优化,得到裂解产品总收益最大的LPG原料组成分布和操作条件,优化后的产品总效益为23.5524万元h~(-1)。
With the increasingly diversified raw materials for ethylene cracking, the proportion of liquefied petroleum gas (LPG) is gradually increased. However, there are few studies on the modeling and optimization of the cracking process using LPG as the raw material. Therefore, Process modeling and optimization research. The cracking mechanism was used to simulate the cracking process of LPG. The process took into account both the composition changes of cracking raw materials and the changes of operating conditions on the yield of key cracking products. Then, according to the experimental design principle, the appropriate amount of simulation data was obtained. Input, yield of key pyrolysis products is output PSOBP neural network yield model. Through the model verification, the average relative error between the ratio of propylene to ethylene in the industrial site and the ratio of propylene to ethylene predicted by this model is only 2.142%. Based on this model, the composition and operating conditions of LPG raw materials were optimized by using PSO algorithm to get the composition distribution and operating conditions of the LPG raw materials with the highest total yield of cracked products. The total benefit of the optimized products was 235.524 million h ~ (-1).