基于NSGA-HVC算法的含高熔点原油单管道炼油系统短期调度优化  

Short⁃term scheduling optimization of a single pipeline refining system with high melting point crude oil using a non⁃dominated sorting genetic algorithm based on hypervolume contribution (NSGA−HVC)

作　　者：侯艳[1] 牛聪 滕少华[1] 朱清华[1] HOU Yan;NIU Cong;TENG ShaoHua;ZHU QingHua(School of Computer Science and Technology,Guangdong University of Technology,Guangzhou 510006,China)

机构地区：[1]广东工业大学计算机学院,广州510006

出　　处：《北京化工大学学报（自然科学版）》2024年第6期28-40,共13页Journal of Beijing University of Chemical Technology(Natural Science Edition)

基　　金：广东省重点领域研发计划项目(2020B010166006);国家自然科学基金(61972102)。

摘　　要：为了解决含高熔点原油的单管道短期调度优化问题,对调度过程中出现的目标函数和约束条件建立数学模型,并采用基于超体积贡献(HVC)的进化算法优化求解该问题。通过种群聚类获得唯一目标值集合,然后计算HVC,并根据目标价值优化参考点的选取,提出改进的基于HVC的非支配排序遗传算法(NSGA-HVC)对含高熔点原油的单管道短期调度问题进行多目标优化求解。实例分析表明,与其他算法相比,NSGA-HVC算法的求解性能更优,超体积(HV)指标高于其他算法10%左右,所得解集表现出更好的多样性和收敛性。将NSGA-HVC算法得到的调度结果与已有文献进行对比,结果表明所提算法在不同目标上的优化效果提升了3.1%~24.1%,整体上调度成本显著降低。In order to solve the short-term scheduling optimization problem for a single pipeline containing high melting point crude oil, mathematical models have been established for the objective function and constraint condi-tions in the scheduling process, and an evolutionary algorithm based on hypervolume contribution (HVC) was used to solve the problem. The unique objective value set was obtained by population clustering, and the HVC was then calculated. The selection of reference points was optimized according to the objective value, and an improved non-dominated sorting genetic algorithm based on HVC (NSGA‒HVC) was proposed to solve the multi-objective optimi-zation of the short-term scheduling problem for a single pipeline containing high melting point crude oil. The analy-sis shows that compared with other algorithms, the NSGA‒HVC algorithm has better solution performance, the hypervolume (HV) indicator is about 10% higher than other algorithms, and the obtained solution set shows better diversity and convergence. The scheduling results obtained using the NSGA‒HVC algorithm have been compared with the existing literature. The results show that the optimization effect of the proposed algorithm for different objectives is improved by 3. 1%‒24. 1%, and the overall scheduling cost is significantly reduced.

关 键 词：高熔点原油 NSGA-HVC算法 短期调度 超体积贡献 多目标优化 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]