关键词: 产量, 产量

Abstract: Evapotranspiration (ET) is one of the main components of the hydrologic cycle. Accurate estimation of ET is essential for studies such as water management and irrigation system design. In this study, a hybrid model that integrated Nonlinear Principal Component Analysis (NLPCA) method with the radial basis function (RBF) neural network (NLPCA-RBF) was used for estimating Tomato ET. On the premise that not only the integrity of the meteorological information can be guaranteed but the correlation among different factors can be eliminated, NLPCA was applied to simplify the seven main meteorological factors which related to the ET into three principal components .Then used these components as input and the measured ET as output target, therefore, the network was created and verified by parts of the data were not used in design of the network. By comparing with RBF neural network, the results showed that NLPCA-RBF network model can well reflected the relationship between meteorological factors and evapotranspiration and got more accurate result.