NBA-MEPP模型在区域水资源安全评价中的应用

利用新型蝙蝠算法(NBA)优化最大熵投影寻踪(MEPP)模型的最佳投影方向,提出NBA-MEPP水资源安全评价模型,并构建了蝙蝠算法(BA)-MEPP、人工蜂群(ABC)算法-MEPP和粒子群优化(PSO)算法-MEPP作为对比模型;从水资源禀赋、水量安全、效率安全以及水质安全4个方面提出了水资源安全评价指标体系和分级标准,并以云南省16个州市的水资源安全评价为例进行实例研究。研究结果表明,NBA算法寻优MEPP目标函数获得的最优值、最劣值、平均值和标准差均要优于BA、ABC和PSO算法,具有较好的全局极值寻优能力。NBA-MEPP模型对德宏州的水资源安全评价为"安全",对丽江市的评价为"临界安全",对其他14个州市的评价为"较安全"。NBA-MEPP模型对16个州市的水资源安全评价结果与BA-MEPP模型的结果相同,但是在排序上存在着差异;与ABC-MEPP、PSO-MEPP模型在评价结果及排序上均存在差异。提出的模型及评价方法具有通用性,可为相关评价研究提供参考。

Application of New Bat Model-Maximum Entropy Projection Pursuit Model in regional water resources security assessment

The new bat algorithm (NBA) was used to improve the optimal projection direction of Maximum Entropy Projection Pursuit (MEPP) model, and the NBA-MEPP water resources security assessment model was proposed. Bat Algorithm (BA) -MEPP, Artificial Bee Colony (ABC) algorithm-MEPP and Particle Swarm Optimization (PSO) Algorithm-MEPP were constructed as the contrast models. The water resources security assessment index system and grading standard were put forward from four aspects of water resources endowment, water quantity security, efficiency security, water quality security. The water resources security assessment in 16 cities or prefectures of Yunnan Province are carried out. The results show that the optimal, worst, average values and the standard deviation of the NBA-MEPP algorithm are superior to those of BA, ABC and PSO algorithms and the NBA-MEPP algorithm has the best global optimal value optimization capability. The NBA-MEPP model assesses the water resources in Dehong as "safe", "critically safe" in Lijiang and "relatively safe" in the other 14 cities or prefectures. The result of NBA-MEPP model is the same with that of BA-MEPP model, but the rankings are different. The result and ranking of NBA-MEPP model are different from those of ABC-MEPP model and PSO-MEPP model. The model and assessment method proposed in this paper are generally applicable and can provide reference for related assessment research.