With the prerequisite of a specific transportation plan, a reasonable formulation for the pump start-up scheme of pump stations along a multiproduct pipeline can ensure a safe and economical operation condition with low energy consump-tion. Based on the further study on the multiproduct pipelines batch transportation, in this paper we consider the power cost of pump stations in the multiproduct pipeline system as the objective function, and a mathematical model aiming to optimize the pump start-up scheme for large-scale multiproduct pipelines is established. The model is based on the discrete-time meth-od. The dynamic programming algorithm and the genetic algorithm are together utilized to solve the optimization issue. The south-west multiproduct pipeline is taken as a real case study, and the calculation results solved by the hybrid two algorithms are compared with those by each of the two algorithms. The results show that the dynamic programming algorithm should be used during the key time, and the ant colony algorithm has better to be used in other cases. In comparison with the previous methods to satisfy the economy of the pump start-up scheme, the computational efficiency of the hybrid method is dramatically enhanced, which can guarantee the stable and efficient operation of large-scale multiproduct pipelines and meet the in-situ process requirements.%在输送计划一定的情况下合理制定成品油管道沿线各个泵站的开泵方案,可以确保成品油管道安全、经济地运行.在对成品油管道顺序输送问题深入分析的基础上,以沿线泵站动力费用总和最少为目标函数,建立成品油管道调度的优化开泵方案数学模型,模型采用离散时间表达.采用蚁群算法与动态规划算法混合求解模型,以西南成品油管道为例进行模拟计算,并将两种算法混合求解的计算结果与采用动态规划算法和遗传算法单独求解结果进行对比.结果表明:输油计划中的关键时刻采用动态规划算法求解,非关键时刻采用蚁群算法求解.相比于前人所采用的算法,在保证所求开泵方案经济性的基础上,大幅度提高了求解计算效率,从而确保大型成品油管道稳定、高效地运行,符合现场工艺要求.
展开▼
