Existing typical scenarios extraction methods fail to fully consider the coupling relationship between uncertain factors, and traditional clustering methods perform poorly on high-dimensional data sets, and the extracted typical scenes cannot well reflect the characteristics of the original data. In view of this situation, on the basis of improving the kernel function, combine Bayesian formula and multivariate Gaussian distribution, use Gaussian Process Regression (GPR) to model the coupling relationship of multiple uncertain factors in the power system, and generate simulating data ;Use the time sequence segmentation typical scenarios extraction method, divide the total scheduling interval into several sub-intervals and perform central point clustering respectively, obtain the weighted typical scenarios of the sub-intervals and connect them with the Cartesian product to generate the typical scenario set of the entire scheduling interval; Evaluate the extraction effect of typical scenarios on the basis of Earth Mover's Distance (EMD) method; The example verifies that the extracted typical scenarios can better retain the probability distribution characteristics of the original basic scenario set, and fully reflect the coupling between different uncertain factors in the original data set and proves that the typical scenarios extracted by the method can better reflect the characteristics of the original data.