论文部分内容阅读
在化工过程合成中,人们在确定研究系统的最大超结构后,通常采用混合整数非线性规划模型将其表达,而后通过计算机对该模型求解,从而找到最佳的流程结构。然而,近年来出现了1种新的求解过程,称为加速分支定界法(ABB),是在最大结构已知的基础上,采用分支定界法进行求解的思路。该算法克服了传统方法在处理整型变量时出现的麻烦,不需要建立复杂的混合整数非线性规划模型,就可以实现计算机自动寻找最优的过程流程,为快速有效地求解化工过程综合优化问题提供了1种新的途径。本文对分支定界法与加速分支定界法进行了详细比较,证实了ABB算法在实现自动寻找最优流程结构的合理性与可靠性。最后,以生化法制备丁醇、乙醇和丙酮的下游分离提纯为实例,研究了ABB算法在过程优化中的应用。结果表明,该算法克服了传统方法在处理整型变量时出现的麻烦,是1种快速有效地求解化工过程综合优化问题的新途径。
In the process of chemical process synthesis, people determine the maximal superstructure of the system and then express it by using mixed integer nonlinear programming model, and then solve the model by computer to find the best process structure. However, in recent years, a new kind of solving process has emerged, called Acceleration Branch and Bound Method (ABB), which is based on the known maximum structure and adopts branch and bound method to solve the problem. The algorithm overcomes the troubles of the traditional method when dealing with integer variables, and does not need to build a complex mixed integer nonlinear programming model, so that the computer can find the optimal process flow automatically. In order to solve the chemical process optimization problem quickly and effectively, Provides a new way. In this paper, a detailed comparison of branch-and-bound method and accelerated branch-and-bound method is carried out, which confirms the rationality and reliability of ABB algorithm in finding the optimal flow structure automatically. Finally, the separation and purification of butanol, ethanol and acetone by biochemical method were taken as examples, and the application of ABB algorithm in process optimization was studied. The results show that the algorithm overcomes the troubles of the traditional method when dealing with integer variables and is a new way to solve the chemical process optimization problem quickly and effectively.