The invention discloses an adaptive linearized probabilistic power flow calculation method including high-proportion wind power connected to the grid, which is used to solve the power system power flow calculation considering high-proportion wind power and strong load random disturbance. The present invention first adopts the Gaussian mixture model to uniformly describe the randomness of the source-load intensity, and accurately establishes the source-load uncertainty model. Then, a stochastic simulation generates a sample of input variable dependencies. Then, the power flow equation is adaptively linearized, and the iterative algorithm is used to realize the adaptive multi-area division of the random fluctuation range, and the linearization is performed within the divided area to reduce the global linearization error of the power flow, and the adaptive linearization semi-invariant method is used The analytical method is used to obtain the semi-invariants of each order of state variables after regional integration and reorganization. Finally, the C-type Gram-Charlier series is used to fit the probability distribution of state variables. The invention can effectively deal with the randomness and correlation of source-load intensity, obtain more accurate power flow distribution, has the advantages of accurate results and convenient implementation, and is suitable for probabilistic risk analysis and evaluation under high-ratio wind power grid-connected.