基于NSGA-Ⅲ算法的合流管网末端截流调蓄设施优化设计

OPTIMIZATION OF DESIGN OF TERMINAL FLOW INTERCEPTION AND STORAGE FACILITIES OF COMBINED DRAINAGE SYSTEM BASED ON NSGA-Ⅲ ALGORITHM

在城市排水系统中,截流调蓄设施是控制合流制溢流(CSO)的一项重要措施。通过将暴雨管理模型(SWMM)与多目标遗传算法(NSGA-Ⅲ)相结合,对截流调蓄设施优化设计问题进行求解。该方法以截流效率、建设成本、水泵启停次数作为模型的优化目标,实现了截流调蓄设施规模的多目标优化。通过对SWMM进行面向对象的重新编码,实现了多线程计算和SWMM与NSGA-Ⅲ模块之间的快速数据交换,避免了频繁的文件操作,求解速率可提升至单线程计算的16倍。利用该方法对武汉市某合流制管网末端截流调蓄设施进行了优化设计,结果表明:该方法所确定的优化设计方案的建设成本可降低为原设计方案的60%,且在截流效率、调蓄池容积等方面更具优势。

The interception facility is an important and frequently-used measure for combined sewer overflows(CSOs) control in city-scale drainage systems. By combining the stormwater management model(SWMM) with the multi-objective genetic algorithm(NSGA-Ⅲ), the optimal design problems of interception and storage facilities were solved. The method took the interception efficiency, construction cost, and the number of pump startup/shutoff times as the optimization objectives of the model, and achieved multi-objective optimization of the interception and storage facility scale. Through object-oriented recoding of SWMM, multi-threaded calculation and fast data exchange between SWMM and NSGA-Ⅲ modules were realized, frequent file operations were avoided, and the solution efficiency was increased to 16 times that of single-threaded calculation. The method was used to optimize the design of terminal flow interception and storage facilities for a combined drainage system in Wuhan. The results indicated the construction cost of the optimized design was reduced to 60% of the original design, and it had more advantages in the aspects of interception efficiency and storage pool volume.