The invention relates to a new energy uncertainty planning method considering typical scene tolerance, and the method comprises the following steps: S1, carrying out the calculation according to the wind speed and illumination intensity data, and obtaining an RDG daily output curve sample; S2, randomly selecting a curve from the sample as a first curve in the initial center; S3, constructing an initial center; S4, calculating Euclidean distances from the remaining curves to K clustering centers , dividing the remaining curves into the corresponding clustering centers, and generating a K-classcurve set; S5, updating K clustering center curves; S6, judging whether a convergence condition is met or not; S7, adopting different K value clusters, calculating contour parameters of correspondingresults, and selecting an optimal cluster K value; S8, constructing a planning objective function, and establishing constraint conditions of the objective function; and S9, obtaining an optimal planning scheme of the renewable distributed energy by using a multi-objective optimization algorithm. The method guarantees that the planning result is suitable for as many operation scenes as possible.