论文部分内容阅读
在模块环境(Aspen Plus)下,建立了基于多目标遗传算法NSGA-Ⅱ求解多目标优化问题的系统结构,并对含循环物流的连续过程废料最小化问题进行求解。在求解过程中遗传算法需要反复调用流程模拟,而流程中循环物流的迭代收敛使优化计算效率较低。为减少流程迭代次数本文提出2个加速策略:一是变收敛精度策略,在优化计算初始阶段,使流程在较低精度下收敛,快速淘汰劣点,随着优化的进行,将流程收敛精度逐步提高,得到高质量的非劣解;二是循环流初值策略,利用已有的计算值,回归决策变量与循环流变量的对应关系,改善循环流初值。实例结果表明,加速策略减少了一半左右的流程迭代次数,效率提高50%,本文提出的求解多目标问题的方法能方便地得到问题的Pareto最优解集,可应用于一般连续化工过程的多目标优化。
Under the Aspen Plus environment, a system structure based on multi-objective genetic algorithm NSGA-Ⅱ for solving multi-objective optimization problems was established and the waste minimization problem of continuous process with cyclic logistics was solved. In the process of solving, the genetic algorithm needs to call the process simulation repeatedly, and the iterative convergence of the recycling logistics in the process makes the optimization calculation less efficient. In order to reduce the number of iterations of the process, two acceleration strategies are proposed in this paper. One is to change the accuracy of the convergence strategy. During the initial stage of optimization, the process converges at a lower precision and the bad points are eliminated quickly. As the optimization progresses, Improve the quality of non-inferior solution; the other is the initial circulation strategy, the use of existing calculations, the regression relationship between decision variables and circulation variables to improve the initial value of the circulation. The experimental results show that the speed-up strategy reduces the number of iterations by about half and the efficiency increases by 50%. The proposed method for solving multi-objective problems can easily obtain the Pareto optimal solution set of the problem, and can be applied to more general continuous chemical processes Goal optimization.