基于WDO-PP模型的区域水资源系统与经济社会生态系统协调度评价

Coordination degree evaluation of regional water resources system and economic social ecosystem based on WDO-PP model

以云南省16个州市为例进行实例研究,提出风力驱动优化(WDO)算法-投影寻踪(PP)水资源系统与经济社会生态系统协调度评价模型,构建布谷鸟搜索(CS)算法、差分进化(DE)算法和粒子群优化(PSO)算法优化的PP模型作对比,进行水资源系统与经济社会生态系统协调度评价。从水资源、经济、社会和生态系统遴选20个指标构建区域水资源系统与经济社会生态系统协调度评价指标体系,并基于指标系列均值及标准差构造"绝对协调"~"极不协调"6个等级的水资源系统与经济社会生态系统协调度评价标准,应用WDO-PP、CS-PP、DE-PP和PSO-PP模型对实例协调度进行评价及分析。结果表明:DWO算法优化PP模型获得的适应度值和最佳投影方向均优于CS、DE和PSO算法,具有较好的求解精度和全局极值能力。WDO-PP模型对昆明、玉溪两市以及怒江、德宏、西双版纳、迪庆4州的水资源系统与经济社会生态系统协调度评价为"协调",其余州市协调度评价为"基本协调"。WDO-PP模型对实例的评价结果与CS-PP、DE-PP模型相同,但在排序上存在差异;与PSO-PP模型的评价结果及排序上均存在差异,表明智能算法极值寻优能力的强弱决定了评价精度的高低。

The wind-driven optimization（ WDO） algorithm-projection pursuit（ PP） water systems and coordination of economic and social ecosystem evaluation model was proposed,and the cuckoo search（ CS） algorithm,differential evolution（ DE） algorithm and particle swarm optimization（ PSO） were build an compared taking 16 cities in Yunnan Province as examples to study the water resources system and economic and social eco-system coordination degree. From the water resources,economic,social and ecological systems,20 indicators were selected to build coordination degree evaluation index system between regional water resources system and economic and social ecosystem. Based on the mean and standard deviation of index series,6 levels of coordination evaluation criteria,from ＂absolute coordination＂to ＂very inconsistent＂,were established. The WDO-PP,CS-PP,DE-PP and PSO-PP models were applied to evaluate and analyze the coordination degree of the examples. The results showed that,the fitness value and the best projection direction of the PP model were better than those of the CS,DE and PSO algorithms,and had good solution accuracy and global extremum ability. The WDO-PP model evaluated the coordination degrees of water resources system and economic and social ecosystem of Kunming,Yuxi,Nujiang,Dehong,Xishuangbanna and Diqing as ＂coordinated＂,and the coordination degrees of other cities were ＂basicly coordinatied＂. The results of WDO-PP model were the same asthose of CS-PP and DE-PP models,but there were differences in the rankings. The differences between PSO-PP model and PSO-PP model indicated that the strength of the extreme value optimization ability of intelligent algorithms determined the level of accuracy.