基于卷积神经网络的水资源配置评价研究

Research on Water Resources Allocation Evaluation Based on Convolutional Neural Network

由于区域水资源配置效果的模糊性极强,为准确评价水资源配置方案的科学性,基于BP神经网络构建水资源配置评价模型。利用LM算法自适应修正BP神经网络权值,采用改进萤火虫算法确定网络模型的初始权值与阈值,由此构建高性能的BP神经网络水资源配置评价模型。以A市水资源配置方案为对象展开仿真评价测试,结果显示:BP神经网络模型评价结果与原有的灰色关联法评价结果趋同,基本可以准确评价水资源配置方案的合理性,对缓解社会水资源短缺有所贡献。

Due to the strong ambiguity of regional water resources allocation effects,in order to accurately evaluate the scientific nature of water resources allocation plans,a water resources allocation evaluation model is constructed based on BP neural network.The LM algorithm is used to adaptively modify the weights of the BP neural network,and the initial weights and thresholds of the network model are determined based on the improved firefly algorithm,thereby to construct a high-performance BP neural network water resources allocation evaluation model.Taking the water resources allocation scheme of City A as the object to carry out the simulation evaluation test,the results show that the evaluation results of the BP neural network model are similar to the evaluation results of the original grey relational method,and the rationality of the water resources allocation scheme can basically be accurately evaluated and resource scarcity can be improved.