城市水资源可持续利用评价的改进BPNN模型——以湖北省为例

Improved BP Neural Network Model of Urban Water Resources Sustainable Utilization Evaluation——Hubei Province as a Case Study

针对水资源评价的复杂非线性,提出城市水资源可持续利用评价的改进BPNN模型。利用基于依赖度和属性重要度的粗集约简算法提取城市水资源评价关键指标,再利用多层BP模式识别网络对数据进行训练实现智能推理和评价,对2012年湖北省内武汉等七大城市水资源可持续利用评价横向比较,用Matlab7仿真结果验证该模型的合理性和有效性。

In view of the complicated nonlinear problems in water resources evaluation, the improved BPNN model of urban water resources sustainable utilization evaluation is proposed. Based on dependency and attribute importance of attribute reduction algorithm, the key indicators of the urban water resources evaluation are extracted, then multi - layer BP network pattern recognition is used to evaluate training intelligent reasoning and data. Matlab7 simulation results on sustainable utilization of water resources in seven major cities about Wuhan etc. of Hubei Province in 2012 validate the rationality and validity of the model.