摘要
为了降低管道运行能耗,针对成品油管道开泵方案的优化求解方法主要包括动态规划算法(Dynamic Programming,DP)和基本遗传算法(Simple Genetic Algorithm,SGA),其存在管道较长时计算效率低及求解结果最优性不高等问题。在对成品油管道运行过程中开泵方案优化问题深入分析以及对并行计算技术深入理解的基础上,应用模拟退火遗传算法粗粒度模型(Coarse-Grained Simulated Annealing-Genetic Algorithm,CGSAGA)进行较大规模的成品油管道开泵方案优化研究。以某实际运行的成品油管道为例进行计算,结果表明:CGSAGA的求解效率与结果最优性均优于传统串行算法,因而为成品油管道开泵方案优化快速、准确制定提供了有效的途径。
The energy consumption of pipeline operation can be reduced greatly by optimizing the pump start-up scheme of products pipelines. When the Dynamic Programming(DP) algorithm is used for long pipelines, its calculation efficiency is low. The Simple Genetic Algorithm(SGA) is low in the optimality of solution results. In this paper, the CoarseGrained Simulated Annealing-Genetic Algorithm(CGSAGA) was applied to optimize the pump start-up scheme of largescale products pipelines, after the issues related with the optimization of pump start-up schemes during the operation of products pipelines were analyzed thoroughly and the parallel computing technology was understood deeply. To verify the effectiveness of the algorithms, an actual products pipeline was taken as an example for calculation. It is shown that the CGSAGA is superior to the traditional serial algorithm in terms of calculation efficiency and solution optimality. It provides the effective way to determine the optimization of pump start-up scheme of products pipelines rapidly and accurately.
出处
《油气储运》
CAS
北大核心
2018年第4期395-402,共8页
Oil & Gas Storage and Transportation
基金
国家自然科学基金资助项目"成品油管道批次输送过程中的复杂传热传质机理研究"
51474228
关键词
成品油管道
模拟退火遗传算法
粗粒度模型
并行计算
开泵方案优化
products pipeline
simulated annealing-genetic algorithm
coarse-grained algorithm
parallel computing
pump start-up scheme optimization
