基于主成分分析的支持向量机需水预测模型及其应用

需水预测是水资源优化配置、水资源规划和水资源管理的重要依据,其预测精度受到众多因素的影响,且实际用水量数据时间系列较短,制约了传统预测方法的应用。利用支持向量机在对小样本学习的基础上对其他样本进行快速、准确的拟合预测的特点,采用主成分分析与支持向量机相结合的方法,首先利用主成分分析法筛选需水量的主要影响因子,然后将其作为输入样本,对支持向量机模型进行训练和检验,寻找最优模型,并将该方法应用于洛阳市需水预测。结果表明,该模型预测结果平均相对误差为-0.83%,预测精度较高,可作为训练样本较少情况下的一种需水预测方法。

Application of Support Vector Machines Based on Principal Component Analysis in Water Demand Prediction

Water demand prediction is an important basis for optimal allocation of water resources,water resources planning and management.Its accuracy of predicting is affected by many factors,and time series of the actual water consumption data is relatively short,affecting the application of traditional forecasting methods.The support vector machine can fit the other samples quickly and accurately on the basis of learning in small samples,so the text uses the method of combining principal component analysis with support vector machines.At first principal component analysis method is used to screen the major water demand impact factor,and then the support vector machine model is tested as an input sample,to find the optimal model.The method is used to predict water demand in Luoyang City.The results show that an average relative error predicted by the model is-0.83%,prediction accuracy is high,so it can be used as a method for forecasting water demand in the case of training samples less.