多级灌溉渠系配水优化编组模型与算法研究

Optimal water allocation marshalling model of multilevel irrigation canal system and model Solution

合理安排各级渠道的配水时间和流量,实现灌溉渠系大流量短历时输配水,减小整个渠系的渗水损失,是目前灌区管理中亟需解决的技术问题。该文基于大系统分解协调的思想,以陕西省冯家山灌区为例,建立了一种解决多级渠系配水优化编组问题的分解协调模型。针对模型中存在多约束条件和大搜索空间问题,设计了基于自适应遗传算法模型求解方法,并比较分析了用该文方法与传统方法确定的配水编组方案性能,结果表明该文确定的算法能实现多约束条件下模型的稳定快速求解,确定的优化配水方案下级渠道配水过程搭配合理,上级渠道配水流量较为均匀且总输水损失小等优点,能解决多级渠系配水优化编组问题,可满足灌区管理需要。

The reasonable arrangement of water allocation time and discharge of each canals, realization of water allocation of large discharge and short time about irrigation canals system and reduction of water transportation loss in irrigation district management are pressing technological problems. Traditional method based on the two-level canal system, can not optimize water allocation marshaling of multilevel irrigation canal system. To solve the problems a new model based on the idea of decomposition and coordination is established. As to many restriction conditions and large searching space problem in the model, solution method based on self-adaptive genetic algorithm is designed, and comparison between the optimal water allocation and the traditional method shows that there are more advantages in the method of this paper than the traditional method. For example, it can analyze the model with multiple restriction conditions quickly and stably using the optimal water allocation program. The water in the lateral channels is allocated more reasonably, and the superior water flow is more uniform, and the total water transportation loss is the lowest. The water delivery scheduling problems of multilevel optimal irrigation canal system can be solved, which meet the requirement of irrigation management.