Research On Distributionally Robust Optimization For Power System Dispatch Considering Moment Uncertainty Of Demand Response

The implementing of demand response programs can improve the economic and security of the power system,because the actual response of the users affected by many factors,the demand response of the user-side contain uncertainties.Therefore,further research of uncertainties of demand response in optimal scheduling is important for the safety and economic operation of the power system.On the demand side,due to the variability of external environment and complexity of the way of using electricity,probability distribution of describing the randomness of the response load is still uncertain,the response load shows double uncertainties of random and rough.So introducing distributional robust optimization under moment uncertainty(DRO-MU),using uncertainty set that constituted by parameters of moments to describe its double uncertainties.The minimized cost of maximization electricity expectation is set as the objective function;considering the convexity of functions,new constraints is that the minimum conditional expectation of positive and negative spinning reserve and line security domain constraint function is not lower than set parameters,building the optimization model and converting this NP-hard problem mentioned above into semi definite programming by dual principle.The simulation results show that the total system cost increased with the increasing if uncertainty set range;with the improving of the confidence level,the system will reduce the economy in order to improve security;by the consideration of double uncertainties for the variable,the scheme under DRO-MU method has lower economy and higher security than CVaR Method.Demand response is implemented under different price signals or compensation policy,the degree of uncertainty for the load is related to the level of compensation.Therefore,proceed from the demand elasticity between demand and load,improving the uncertain set by introducing an amplification factor,avoiding the possible situation of poor economic caused by the overly conservative parameters of the uncertain set,DRO-MU optimization model considering the uncertainty of demand response and response reliability is established.The simulation results show that higher level of reliability causes relatively lower total system cost,so the power company can determine the optimal solution after weighing the reliability and economic of scheduling and compensation programs under different demand response policies and compensation standards;for the users that have different demand elasticity,the corresponding compensation scheme could be adjusted to reach the harmonization of economic and reliability;and the total cost is increasing with larger correlation coefficient between the load nodes.The improved optimization model under DRO-MU method canprovide a more reasonable scheduling scheme and it’s important for the research of economic scheduling.This method can also be used to solve optimization problems containing other uncertainties like prices or new energy resources and so on.