【摘要】 在制定原油一次加工过程详细调度时,往往需要考虑多个优化目标。本文提出一种基于改进的骨干粒子群算法和II代非支配遗传算法协同进化的双种群算法,并通过Pareto差熵控制种群的交流,优化了供油罐使用成本、供油罐的切换成本、管道中原油混合成本以及供油罐罐底混合成本4个目标。通过一个工业实例,将本文算法与现有的几种具有代表性的进化多目标优化算法进行对比,验证本文算法的可行性和有效性。更多还原

【Abstract】 To detail the schedules for crude oil operations, it is necessary to optimize multiple objectives. A dual population algorithm is proposed,being called bare bone particle swarm optimization and NSGA-II algorithm, BPGA, which is based on the improved bare bone particle swarm optimization algorithm(I-BBPSO) and non-dominated sorting genetic algorithm II(NSGA-II). It controls population exchange through Pareto difference entropy, which optimizes all the costs of the charging tanks, switching the charging tanks, mixing crude oil in the pipeline and mixing crude oil in the bottom of the tanks. The proposed algorithm is applied to an industrial example and compared with several representative evolutionary multi-objective algorithms and shows its feasibility and validity.更多还原