基于DP-PSO算法的灌区农业水资源优化配置  

Optimal Allocation of Irrigated Agriculture Water Resources Based on DP-Particle Swarm algorithm

作　　者：刘博[1] 崔远来[1] 尹杰杰[1] 刘方平 郑传举 

机构地区：[1]武汉大学水资源与水电工程科学国家重点实验室,武汉430072 [2]江西省灌溉试验中心站,南昌330201 [3]湖北省樊口电排站管理处,湖北鄂州436001

出　　处：《节水灌溉》2016年第8期117-121,共5页Water Saving Irrigation

基　　金：江西省水利厅重大竞争性项目(KJ201409)

摘　　要：针对引水灌区,考虑灌溉用水总量约束和时段可供水量约束,以单一作物非充分灌溉下灌溉制度优化为第一层,区域多种作物种植结构及水量分配为第二层,分别采用动态规划和粒子群算法求解,建立了基于DP-PSO算法的灌区农业水资源优化配置模型。针对多重约束问题,提出了一种初始化粒子群的方法,并通过粒子速度的动态变化来保证每代粒子都满足约束,提高了算法的收敛速度和收敛精度。以赣抚平原灌区为对象,考虑降雨和水源可供水量不同步的特点,计算了3种降雨频率、多年可供水条件下的农业水资源优化配置方案。结果表明,基于DPPSO算法的农业水资源优化配置模型合理可靠,为引水灌区农业水资源优化提供了一种新的方法。Aiming at diversion irrigation districts, considering the constraints of both total irrigation water and available water supply of each irrigation period, taking the optimal irrigation water allocation {or single crop under non-sufficient irrigation condition as the first level, the optimal water allocation for different crops as the second level, solving by DP and particle swarm optimization algo- rithm separately, an Optimal Allocation of Irrigated Agriculture Water Resources Based on DP-Particle swarm algorithm was estab- lished. In regard to multi-constraints problem, a method of initializing particle swarm was proposed, by changing particle velocity dynamically, particles of each iteration can satisfy the initial constraints and convergence rate and precision of the algorithm were improved. Taking Gan{u Pain Irrigation District as obiect of study, based on the asynchrony of rainfall arid available water supply, the schemes of irrigated agriculture water resources in years of three typical rainfall frequency under multi-available water supply condi- tions were proposed. The results indicated that the model was reasonable and reliable for optimal allocation of irrigated agriculture water resources in diversion irrigation districts.

关 键 词：引水灌区 多重约束 粒子群算法 水资源 优化配置 

分 类 号：S274[农业科学—农业水土工程] TV211.11[农业科学—农业工程]