APP下载

基于粒子群算法的动态多目标优化

2021-08-09李青

粘接 2021年6期

李青

摘 要:针对碳二氢生产中的反应器动态优化问题,目前虽然有多种算法对生产过程进行优化,但大部分只是对单一目标进行求解,提出一种更为灵活的反应器动态求解方法。在该方法中,首先构建碳二氢目标函数,然后采用多目标粒子群算法和分段线性函数参数法结合的方式对目标函数的进行求解,以提高整体搜索能力,得到碳二氢反应器动态优化的最优解。最后,以实际乙烯碳二加氢化工反应过程为例进行实验验证,结果证明,通过该方法进行求解的目标函数无论是在收敛性,还是在优化的平均值等方面,都比SADE-eCD和NSGA-II算法具有优势,说明该算法在反应器动态优化中是切实可行的。

关键词:动态多目标优化;粒子群算法;碳二加氢;骨干粒子群算法

中图分类号:TP301.6 文献标识码:A 文章编号:1001-5922(2021)06-0039-05

Abstract:In view of the dynamic optimization of the reactor in the production of carbon dihydrogen, although there are many algorithms to optimize the production process, most of these algorithms only focus on the optimization of a single objective, and a more flexible method of reactor dynamics is proposed. In this method, the C2H objective function is first constructed, and then the objective function is solved by the combination of multi-objective particle swarm algorithm and piecewise linear function parameter method to improve the overall search ability, and the optimal solution for dynamic optimization of the carbon dihydrogen reactor is obtained. Finally, the actual ethylene carbon two hydrogenation chemical reaction process is used as an example for experimental verification, and the results prove that the objective function solved by this method has advantages over the SADE-eCD and NSGA-II algorithms in terms of convergence and average value of optimization, indicating that the algorithm is feasible in reactor dynamic optimization.

Key words:dynamic multi-objective optimization; particle swarm optimization; C2 hydrogenation; backbone particle swarm optimization algorithm

近年来,随着化学工业的发展,化工过程的动态模拟越来越受重视,分线性等模型也在化工过程建模中普遍存在。但由于这类模型具有不确定性和高维特性等问题,因此求解难度较大。另外,目前的这类研究算法中,主要关注的是单一目标的动态优化,不符合实际化工过程中多目标动态优化。针对这类问题,本文提出具有灵活有效的约束动态多目标骨干粒子群算法,并通過引入分段线性函数参数化方法,提高了该算法的的局部搜索能力,解决了化工过程中反应器动态多目标优化问题。最后,通过将该算法应用到实际乙烯碳二加氢化工反应过程,结果证明,该算法对目标的优化达到了预期目的,也就是说,该算法在实际化工过程中是切实可行的。……

登录APP查看全文