基于PSO-GRA耦合决策的排水沟系统多目标规划模型

A Drainage System Multi-objective Planning Model Based on PSO-GRA Coupling Decision

排水沟系统承担排水蓄涝、环境保护等效能,多目标排水沟系统规划具有十分重要的现实意义。综合考虑工程费用、减污与产量效应构建排水沟系统多目标规划模型,运用粒子群算法(PSO)生成规划方案非劣解集,并耦合灰色关联投影法(GRA)优选非劣解集进行方案决策,将模型应用于上海松江高标准农田稻作区规划排水沟系统。结果显示,稻作区优选规划方案推荐排水历时5 d,农沟与支沟沟深1.3、1.7 m,间距77、270 m,优化水面率7.64%。与传统单目标规划方案相比,设计降水条件下,增加滞涝水量648.5 m^3/hm^2,总氮的单位面积削减能力4.22 kg/hm^2,削减率增加28.1%,水稻相对产量96.2%。该模型能客观权衡多目标的矛盾性进行科学决策。

As the drainage system is critical for drainage storage,environmental protection and so on,the plan of multi-objective drainage system has a very important practical significance.In this paper,a multi-objective planning model of the drainage system is constructed,in which construction cost,pollution reduction,and capacity effect are taken into consideration. The Pareto set of the planning scheme is generated by using particle swarm algorithm（ PSO）,and planning decision is made by coupling grey related projection（ GRA） to optimize Pareto set. This model is used in the drainage system in Shanghai Songjiang High-standard Farmland-rice Cropping Region. The results show that the optimal plan for the rice area recommends drainage duration of 5 days,ditch depth of 1.3 m and 1.7 m,distance of 77 m and 270 m,and optimized water rate of 7. 64%. Compared with traditional single-objective planning scheme under rainfall design criteria,this model increases stagnant water with 648.5 m^3/hm^2,reduces total nitrogen per unit area 4.22 kg/hm^2,the reduction rate of total nitrogen increases by 28.1% and the relative rice production is about 96.2%. This model can objectively weigh the contradiction of multi-objectives and make scientific decisions.