| With the continuous improvement of economic level,human demand for electric energy is increasing,and the power system is gradually developing to the direction of large capacity,multi energy and wide range.The traditional power dispatching mode,which is limited to the generation resources,may cause problems such as the renewable energy can not be reasonably absorbed and the system needs to be continuously expanded,which will further increase the difficulty of system operation and dispatching.With the development of smart grid,the demand side has become an indispensable interactive resource.It is an inevitable requirement and a new challenge for the development of power system to add demand response to power system dispatching and consider the impact of its uncertainty.In addition,the real distribution of uncertain demand response is ambiguous for decision-makers,so how to build an appropriate data-driven demand response uncertainty set in the scheduling model is worth further study.Therefore,this paper studies the power system scheduling problem considering demand response uncertainty.The main contents are as follows:Firstly,on the basis of arranging the application status of demand response and the research status of participating in scheduling,this paper analyzes the mechanism of demand side participating in system scheduling,including clarifying the subject of demand response market and various modes of participating in the demand response market.Respectively,the model of demand response participating in the system optimal dispatching is established.Based on the power demand-price curve,the response model of real-time price project is established and its implementation cost function is constructed.The flexible interruptible load is selected as the representative of incentive demand response to build the response model and its implementation cost function,which paves the way for the later research.Then,the principle of distributionally robust optimization algorithm and its classification application are briefly introduced.The applicability of the method to the power system scheduling problem considering the uncertainty of demand response is analyzed.On this basis,considering the uncertainty of users’ response to the real-time price project,the corresponding ambiguity set based on Wasserstein metric is constructed.A two-stage distributionally robust optimization model for day-ahead dispatch of power system is proposed,and the corresponding dual transformation solution method is proposed.The simulation result shows that the proposed scheduling scheme sacrifices part of the economy in exchange for the safety of the system operation.It can obtain more stable calculation results under the condition of large sample ambiguity and small sample number,and has excellent out-of-sample performance and high reliability.At the same time,changing the confidence of ambiguity set can balance the robustness and economy of distributionally robust scheduling scheme,which makes it more flexible.Finally,the reserve market mechanism of power system is simply analyzed,and the traditional unit reserve and interruptible load reserve are modeled.Under the assumption that the reserve market and the main energy market are jointly optimized,and all the reserves are cleared in the day-ahead market,a two-stage distributionally robust optimization model for the joint dispatching of the main and reserve power systems considering the response uncertainty of interruptible loads is established.Through the example verification,the distributionally robust optimization model can reserve a certain amount of units and interruptible load as reserve capacity in the pre-scheduling stage,and avoid a lot of power imbalance caused by the uncertainty of demand response.Interruptible load,as a reserve resource,can effectively expand the upward reserve capacity of the system,provide a new way for system reserve dispatching,and enhance the economy of system scheduling and flexibility of operation. |
PDF Full Text Request