A Two-Stage Distributionally Robust Unit Commitment Model Considering Multiple Uncertainties

The uncertainties of the output of renewable energy units and the load demand of power system have always been important issues in the research field of power system unit commitment.As the installed proportion of gas-fired units gradually increases in the construction of the renewable power system,the uncertainty of the output limitation of gas-fired units caused by the malfunction risk of gas network also becomes important enough.If the contingency of the uncertain factors such as the output limitation of gas-fired units and the output of renewable energy and the load demand appears simultaneously,the safety of the power system will be seriously threatened.In order to improve the robustness of the unit commitment scheme,the multiple uncertainties should be considered by the dispatching agency during developing the unit commitment scheme.Therefore,it is of great significance to use scientific method to quantify the output limitation of gas-fired units and further study the power system unit commitment model considering multiple uncertainties on the basis of the existing research on uncertain optimization.In this paper,a quantitative analysis method of the uncertainty risk of the output limitation of gas-fired units was proposed firstly.By establishing the malfunction model of natural gas network components,the Monte Carlo method was used to randomly sample the historical statistics on failure rates of natural gas network components to generate the natural gas network malfunction scenarios.Then,based on the dispatching strategy of "reducing others to ensure the residential load" of the natural gas dispatching agency,the optimal reduction model of natural gas load was proposed to calculate the natural gas flow of the gas network nodes where the gas-fired units are located in each gas network malfunction scenario,and the maximum available output of the gas-fired units under the malfunction scenario was obtained according to the currently frequently-used gas-electricity conversion relationship of the gas-fired units.On this basis,the states probabilities table output limitation of gas-fired units was obtained by the numerical results on the Belgium 20-bus natural gas network,the validity of the proposed analysis method for the states probabilities output limitation of gas-fired units was verified,and the uncertainty of the output limitation of gas-fired unitswas characterized by "strong correlation".Further,the paper introduced and analyzed the characteristics of three types of probability ambiguity set frequently-used in distributionally robust optimization method.Since the three types of probability ambiguity set constitution can effectively describe the uncertainties of the output of renewable energy units and the load demand,this paper mainly based on the "strong correlation" of the gas-fired units’ output limitation,selected the probability ambiguity set based on multiple discrete scenarios to describe the uncertainties of the output limitation of gas-fired units and the output of renewable energy units and the load demand.On this basis,a two-stage distributionally robust unit commitment model considering multiple uncertainties was proposed,and the specific objective function and constraints of the model were given.The proposed model was solved by using duality-free column and constraint generation algorithm.The results demonstrated that the distributionally robust unit commitment model proposed in this paper can effectively deal with the unit commitment problem considering multiple uncertainties,the robustness of the unit commitment result was significantly improved with the economy decreased slightly,and the model has a good regulation ability of conservative degree,can flexibly adapt to different application scenarios according to the number of the historical samples and the risk preference of dispatchers.