Studies On The Model And Its Algorithm Of Distributional Robust Optimization Under Moment Uncertainty Of Dispatch For Wind Integrated System

As relies on the wind,wind power has greater uncertainty and seriously affects the economic and reliable operation of the power grid when its large-scale connected to the grid.Therefore,research in-depth of optimal dispatch model,which can deal with uncertainty problem,have great practical significance in wind integrated system.The stochastic dispatching model for wind integrated system can only handle the wind power sequence,which have certain probability distribution.However,the result of ADF test shows that the probability distribution of wind power sequence has moments uncertainty.So,the model of distributional robust optimization under moment uncertainty of dispatch for wind integrated system is established by a new model named Distributional Robust Optimization under Moment Uncertainty(DRO-MU)to solve this problem.In modeling process,the moments uncertainty sets containing the expectation and covariance matrices of the wind are used to describe the uncertainty of the probability distribution of wind power.Objective function is minimizing the maximum expected value of generation cost in the scope of its uncertainty sets uncertainty set.Considering the mathematical convexity,the constraints are using the inequality which maximum conditional expectation of spinning reserve and line security domain is not greater than the setup parameter.In the solving process,the model of DRO-MU of dispatch for wind integrated system is a NP hard problem,so this model is transformed into semi-definite programming by using Lagrangian duality principle.Simulation results show that the dispatch cost increase with the increase of the uncertain set and correlation coefficient additional.The different confidence levels result different dispatch schedule,which emphasize on different security demand.The dispatch cost considering of uncertainty of the moments is higher the calculation of the Conditional Value-at-Risk method under a certain distribution,but the scheme is more secure.The DRO-MU dispatching model based on stochastic dispatch models,and considering the uncertainty of the wind’s moments,like expectation and covariance.It is a new mentality and model of solve the moment uncertainty problem in the wind integrated system.This model is based on single generator controlling frequency system,leading to emerge the phenomenon that slack generator overflowed its adjustable range,and the expression of nearly analytic function of power flow is not accurate.But the ratio of frequency modulation is not to make reasonable decision in the dispatch model of multiple AGC units system.So,adopting DRO-MU dispatch model to find a decision of ratio of frequency modulation in multiple AGC units system.Due to it is a complex semi-definite programming,which con-tains higher-order decision variables after dual transformation,there is no effective algorithm to solve this model.Thus,that an improved Generalized Benders Decomposition algorithm(GDB)ameliorated by the minimum eigenvalue optimization is used to solve this problem.Simulation results show that the improved GDB algorithm will converge and used to solve our model.The dispatch cost increase with the increase of the uncertain set and correlation coefficient additional.The dispatch cost considering of decision of the ratio of frequency regulation is lesser the calculation of the given ratio,and more security.Decision of the ratio of frequency regulation of AGC units can provide a more reasonable deviation to power allocation scheme in electric dispatching.Has the vital significance to the economic dispatching of the wind integrated system.