摘要
针对水资源评价的复杂性不确定性等问题,提出了一种粗集与BP神经网络相结合的城市水资源可持续利用评价模型。利用粗糙集属性约简算法对城市水资源评价指标约简,找出主要评价指标,以简化BP网络的输入层;利用BP神经网络的非线性适应性信息处理能力对评价数据进行量化训练。基于湖北省2004~2010年城市水资源的相关数据,利用Matlab仿真平台进行评价,实证结果验证了该改进模型的科学性和有效性,为城市水资源可持续利用评价提供了有效的方法。
In view of the complex uncertainty problem of water resources evaluation, a RS-BP neural network model of urban water resources sustainable utilization evaluation is put forward. Firstly, urban water resources evaluation indexes are reduced by using rough set attribute reduction algorithm in order to simplify the input layer of BP network. And then the evaluation data are trained with the nonlinear adaptive information processing capability of BP neural network. Finally, urban water resources data from 2004 to 2010 in Hubei province are evaluated by using Matlab simulation program. The example results show that the improved model is scientific and effective, which provides an effective way for evaluation of urban water resources sustainable utilization.
出处
《水电能源科学》
北大核心
2014年第6期22-24,32,共4页
Water Resources and Power
基金
湖北省教育厅科研项目(Q20132204)
关键词
属性约简
城市水资源
可持续利用
BP神经网络
attribute reduction
urban water resources
sustainable utilization
BP neural network
