文山州近10年水资源利用效率评价SLC-PP模型及应用

SLC-PP model and its application to evaluation of water use efficiency in Wenshan Prefecture in last ten years

为科学、客观地评价区域水资源利用效率,提出足球联赛竞争(SLC)-投影寻踪(PP)水资源利用效率评价模型,以文山州2006—2015年水资源利用效率评价为例进行实例分析。通过6个典型高维测试函数对SLC算法进行验证,并将仿真结果与教学优化(TLBO)算法、灰狼优化(GWO)算法、差分进化(DE)算法、混合蛙跳算法(SFLA)、粒子群优化(PSO)算法和帝国竞争算法(ICA)的寻优结果进行比较。从综合、工业、农业、生活和生态环境用水5个方面选取23个指标构建水资源利用效率评价指标体系,应用SLC-PP模型,并基于投影系列均值z及标准差σ构造低水平、较低水平、中等水平、较高水平和高水平5个等级的水资源利用效率评价标准对实例进行评价,并分析近10年的水资源利用效率变化趋势。结果表明:(a)无论是单峰还是多峰函数,SLC算法寻优精度优于TLBO和GWO算法,远优于DE、SFLA、PSO和ICA算法,具有寻优精度高、收敛速度快、极值寻优能力强以及收敛稳定性与可靠性好等特点。(b)SLC-PP模型对实例水资源利用效率评价结果均为:2006—2007年处于低水平,2008—2012年处于中等水平,2013—2015年处于较高水平,水资源利用效率随时间呈提升趋势,且提升趋势显著。

For the scientific and objective evaluation of regional water use efficiency, a projection pursuit （PP） model based on the soccer league competition （SLC） algorithm was established. The evaluation of water use efficiency in Wenshan Prefecture during the period from 2006 to 2015 was used as an example. The SLC algorithm was validated using six typical high-dimensional functions, and the simulation results were compared with the optimization results of the teaching-learning-based optimization （TLBO） algorithm, gray wolf optimization （GWO） algorithm, differential evolution （DE） algorithm, shuffled frog-leaping algorithm （SFLA）, particle swarm optimization （PSO） algorithm, and imperial competitive algorithm （ICA）. From five types of water use, including comprehensive water use, industrial water use, agricultural water use, domestic water use, and ecological environmental water use, 23 indices were selected to build an index system for the evaluation of water use efficiency. A water use efficiency evaluation standard with five grades （low level, relatively low level, medium level, relatively high level, and high level） were constructed based on the mean and standard deviation values of the projection series. Based on this standard, the SLC-PP model was used to analyze the trend of water use efficiency in the study area over the past ten years. The results showed the following： First, whether for the optimization of unimodal or multimodal functions, the SLC algorithm performs better than the TLBO and GWO