基于主成分分析的RBF神经网络在需水预测中的应用

发布时间：2018-11-19 14:01

【摘要】：针对需水预测方法大都存在一定的局限性,导致预测值与实际值差别较大的问题,采用主成分分析与径向基函数(RBF)神经网络相结合的方法建立需水量预测模型。首先借助SPSS进行主成分分析,对影响因子进行降维处理,以此减少各影响因子之间原有的多重共线性;其次,选用RBF网络,运用Matlab神经网络工具箱,建立了基于主成分分析的RBF(PCA-RBF)神经网络需水预测模型;并以辽宁西部地区凌源市需水预测为例,对预测模型进行了校核。利用训练好的PCA-RBF神经网络需水预测模型对凌源市2014、2015年的总需水量进行模拟预测,预测结果与实测数据相对误差分别为2.9%、0.4%。这说明该模型可相对全面地模拟需水量变化规律,能够用于半干旱山区和材料相对较少时需水量的精准预测,为水资源规划管理提供了理论依据。
[Abstract]:In view of the limitation of most forecasting methods for water demand, which leads to the big difference between the predicted value and the actual value, the method of combining principal component analysis and radial basis function (RBF) neural network is used to establish the water demand forecasting model. Firstly, the principal component analysis (PCA) of SPSS is used to reduce the dimension of the influence factors, so as to reduce the original multiple collinearity among the factors. Secondly, RBF neural network and Matlab neural network toolbox are used to establish the RBF (PCA-RBF) neural network water demand prediction model based on principal component analysis, and the prediction model is checked by taking Lingyuan city in western Liaoning as an example. Using the trained PCA-RBF neural network water demand prediction model, the total water demand of Lingyuan city in 2014 and 2015 was simulated and forecasted. The relative error between the forecast result and the measured data was 2.9% and 0.4g, respectively. The results show that the model can simulate the variation of water demand relatively comprehensively, and can be used to predict water demand accurately in semi-arid mountain areas and when the materials are relatively small, which provides a theoretical basis for the planning and management of water resources.
【作者单位】：沈阳农业大学水利学院;凌源市水务局;辽宁省水文水资源勘测局;丹东市林业局;
【分类号】：TV213.4

【参考文献】