论文部分内容阅读
热偶精馏是一种新的节能方式,但传统的热偶精馏模拟计算过程繁复,且基于传统模拟过程的优化方法也难以得到良好的可行解。针对这一问题,本文利用人工神经网络和遗传算法的特点,提出将二者结合以改进热偶精馏过程的优化方法,并将该方法应用于丁二烯分离及乙腈回收流程的研究中,先由神经网络建立黑箱数学模型,再由遗传算法优化热偶精馏的操作参数,通过计算结果比较,表明该方法可以迅速得到优化变量和目标函数,而且获得全局最优解。
Thermocouple distillation is a new energy-saving method, but the traditional thermocouple distillation simulation calculation process is complicated, and the optimization method based on the traditional simulation process is difficult to get a good feasible solution. In order to solve this problem, this article proposes the optimization method of combining the two to improve the thermocouple distillation process by using the characteristics of artificial neural network and genetic algorithm. And this method is applied to the research of butadiene separation and acetonitrile recovery process, The mathematical model of the black box is set up by the neural network firstly, and the operating parameters of the thermocouple distillation are optimized by the genetic algorithm. The comparison of the calculation results shows that the method can get the optimization variables and the objective function quickly, and obtain the global optimal solution.