Power system economic dispatching method based on polynomial chaos approximate dynamic programming

The invention discloses an electric power system economic dispatching method based on polynomial chaos approximate dynamic programming, which comprises the following steps: S1, constructing an optimal dispatching model of an electric power system, establishing an objective function about a state variable and a decision variable, introducing uncertain external information, and determining the state of the electric power system by the state variable, the decision variable and the external information; determining a state transfer function; s2, by solving a value function of the optimal scheduling model under the state variable, time period decoupling solving of the multi-time-period power system economic scheduling problem is achieved, and the value function of the optimal scheduling model under the state variable is obtained; s3, constructing a polynomial chaos proxy model about value function distribution, realizing identification of a coefficient to be solved of the polynomial chaos proxy model according to