采用支持向量机(SVM)方法对太阳总辐射值进行了估算,并与传统的线性方法进行了对比分析。对于日太阳总辐射估算,SVM方法的均方根误差为2.27(MJ·m-2)/d,平均百分比误差为22%;而对于月太阳总辐射估算,SVM方法的均方根误差约为21(MJ·m-2)/month,平均百分比误差低于5%。SVM方法提供了一种高精度估算太阳总辐射的新途径。
In this paper,the support vector machine(SVM) method was employed for the global solar radiation estimation and the estimation result was compared with that generated with the traditional linear method.For the daily global solar radiation estimation,the RMSE by the SVM method is 2.27(MJ·m-2)/day and the MAPE is 22%;while for the monthly global solar radiation estimation,the RMSE by the SVM method is about 21(MJ·m-2)/month and the MAPE is less than 5%.The SVM method provides an alternative for global solar radiation estimation with high precision.