基于极限学习机的长江流域水资源开发利用综合评价

为能客观、准确地对长江流域水资源开发利用进行综合评价,利用层次分析法构建了符合长江流域水资源开发利用现状的综合评价指标体系和分级标准,基于极限学习机(ELM)算法原理,构建了ELM水资源开发利用综合评价模型对长江流域及主要水系水资源开发利用进行综合评价,并构建RBF、BP神经网络模型作为对比评价模型。采用随机内插的方法在各评价分级标准阈值间生成训练样本和检验样本,在达到预期评价精度后将模型运用于长江流域水资源开发利用综合评价中。结果表明:ELM水资源开发利用综合评价模型对长江流域及主要水系水资源开发利用综合评价等级为4～8级,处于有潜力至失衡之间,与长江流域各主要水系水资源开发利用现状相符;该模型的评价精度和泛化能力均优于RBF及BP神经网络评价模型,是合理可行和有效的,可应用于长江流域水资源开发利用综合评价,具有参数选择简便、评价精度高、泛化能力强等优点。

Comprehensive evaluation of water resources development and utilization in Yangtze River Basin based on extreme learning machine

In order to accurately and objectively evaluate water resources development and utilization in the Yangtze River Basin,a comprehensive evaluation index system and classification standard of water resources were determined using the analytic hierarchy method.Based on the extreme learning machine(ELM) algorithm,a comprehensive evaluation model for water resources development and utilization was built to evaluate the water resources of the Yangtze River Basin using RBF and BP neural network models as comparing evaluation model.Training samples and testing samples were generated using the random interpolation method to train the ELM comprehensive evaluation model.The results show that the evaluation grade of water resources of the Yangtze River Basin is 4 to 8,which coincides with the water resources reality.Compared with RBF and BP neural network models,the ELM comprehensive evaluation model has better evaluation accuracy and generalization ability,and has the advantages of simplicity,high accuracy,and strong generalization ability.