基于GRNN模型的区域水资源可持续利用评价

利用层次分析法构建了符合丰水地区水资源可持续利用指标体系和评价标准，从水资源条件、水资源开发利用效率、生态环境状况、水资源合理配置和水资源管理能力5个方面提出50个评价指标，并运用GRNN神经网络水资源可持续利用评价模型，对文山州不同规划水平年水资源可持续利用进行了综合评价。结果表明：① 不同规划水平年各评价区域水资源可持续利用评价为2~3级，反映出文山州现状及中、长期水资源处于可持续与基本可持续之间的状况，亦表明所建立的GRNN评价模型和评价方法是合理可行的；② GRNN网络评价模型具有收敛速度快、预测精度高、参数调整少、不易陷入局部极小值等优点，可以更快地预测评价网络，能够满足客观评价区域水资源可持续利用状况的要求。

Evaluation of sustainable utilization of regional water resources based on GRNN neural network model: case of Wenshan Prefecture of Yunnan Province

AHP is used to establish an index system and evaluation standard of water resources sustainable utilization for water-abundant area,and 50 evaluation indexes are put forward from 5 aspects of water resources,water resources development and utilization efficiency,ecological environment,rational allocation of water resources and water management,GRNN neural network evaluation model for sustainable use of water resources is established by the principle of GRNN neural network theory and methods,and the model is used to comprehensively evaluate the water resources sustainable utilization of Wenshan Prefecture in different planning years.The results show that: in different planning years,the sustainable utilization of water resources of Wenshan Prefecture is evaluated as grade 2 to 3,namely,between sustainable utilization and basically sustainable utilization,which reflects the current situation and the mid-long term sustainable utilization of water resources and is consistent with the actual development of the region;GRNN network evaluation model has the characteristics of fast convergence,high prediction accuracy,less parameters adjusting,uneasy to fall into local minimums,and can predict and evaluate more quickly to meet the objective evaluation requirements of regional water resources sustainable utilization.