基于复合多目标方法的灌区水资源优化配置

Optimally Allocating of Water Resources among Irrigation Districts Based on Composite Multiobjective Method

摘要:

针对传统灌区水资源多目标模型求解过程中难以兼顾多维配置要求的问题，基于目标协调度、可持续发展指数、目标实现度3个指标提出了一种复合多目标方法。为验证该方法有效性，以甘肃省黑河中游17个灌区间水资源优化配置为例，以经济效益、社会效益、生态效益为目标函数构建多目标优化模型，并分别使用传统方法与复合多目标方法进行求解。结果表明，使用复合多目标方法所获得的优化配置方案可以将灌溉水利用系数提高5.42%~7.57%。结果验证了传统多目标模型在协调多个冲突目标方面较单目标模型有显著优势，而复合多目标方法所获得的优化方案更能体现决策者对研究区域种植业发展与灌区水资源配置的多元要求。

Abstract:

The optimal allocation of agricultural water resources is an effective way to improve the utilization efficiency of agricultural water resources. The formulation of agricultural water resources allocation scheme needs to consider multiple conflicting objectives, which can be solved by building a multiobjective optimal allocation model. When solving such multiobjective models, the multiobjective model is often transformed as a single objective model with the goal of maximizing target achievement. However, such method can hardly reflect other aspects that the decision-makers are concerned about, such as the synergy degree of different goals and the impact of the scheme on the sustainable development ability. Moreover, setting only one index as the objective function cannot fully reflect the comprehensive impact of multidimensional optimization needs on the optimization scheme. Therefore, a composite multiobjective method was proposed to optimally allocate agricultural water resources based on the three indicators, including coordination degree of multiple objectives, sustainable development index, and target realization degree. In order to verify the effectiveness of the method, composite multiobjective method was applied to a real-world case in the middle reaches of Heihe River in Gansu Province for optimally allocating agricultural water resources among 17 irrigation districts. A multiobjective optimization model with objective functions of economic, social and ecological benefits was developed, and different solving methods were employed to solve the model. The optimization results showed that the agricultural water resources utilization coefficient can be increased by 5.42%~7.57% by using the composite multiobjective method. The results of single objective model were too extreme to be applied to practical water resources allocation comparing with results of multiobjective model. Compared with the multiobjective models which only used one index as the transformation objective function, the optimization scheme obtained by the composite multiobjective method can well reflect the multiple requirements of decision-makers for regional agricultural development and agricultural water resources allocation.