A Distributionally Robust Bilevel Optimal for Low-Carbon Power System Considering Load Uncertainty

Abstract

With the depletion of fossil energy worldwide, new energy power generation represented by photovoltaic, wind power, etc. is being applied, which introduces a nonnegligible supply and demand balance difficulty to the power grid. Relying solely on the regulating capacity of traditional power plants is insufficient, it is urgent to awake demand-side resources and construct a multi objective optimization model that considers the interests of both supply and demand to achieve flexible interaction between power users. To address this problem, a multi objective distributionally robust bilevel optimization model based on KL divergence is proposed, which optimizes system operation cost, carbon emission and unit profit, while resolving the self-scheduling problem under load and price uncertainties. Additionally, the GARCH model is introduced and applied to the Value at Risk theory, which characterizes the fluctuation aggregation and heteroscedasticity of the daily load change of the load aggregator, so that people can grasp the risk more accurately. Subsequently, a novel approach for bilevel programs based on Wolfe duality is employed to transform the bilevel optimization problem into a single level optimization problem, and the nonlinear multi objective problem is directly solved by GUROBI. Finally, the simulation of the case shows that the equivalent DR model can realize efficient allocation and complementary optimization of flexible resources in the market environment and improve the operation economy and stability of the power system.