Research Article

Wind Power Prediction Based on Nonlinear Partial Least Square

Table 2

Analytical results for working condition II.

ModelRMSECVRMSEPLVsHidden nodesWidth

PLS309.3051278.41530.74180.79992--
BP258.7904250.92900.75270.7908-1-
RBF477.7540460.56280.84290.2303-1-
SVM508.0016336.42370.70370.6328--0.6
BPE-NPLS286.3022234.46760.76360.787612-
RBFE-NPLS217.8454207.93190.82170.834125-

The bold means the minimum error or the maximum correlation among models.