Optimal Scheduling Of Active Distribution Networks Considering Temporal And Spatial Correlation

With the rapid integration of renewables,optimal economic dispatch in active distribution networks faces great challenges because of the randomness and volatility of renewable energy and complex distribution network structure.Low prediction accuracy and longer solving time cause that it is difficult to meet the optimal economic dispatch in traditional optimal model.Therefore,it is necessary to study the optimal dispatching model of active distribution networks that takes into account the uncertainty of renewable energy output.The dissertation focuses on the optimal dispatching of active distribution networks.Focusing on the above problems,the major work in this dissertation is summarized as follows:(1)The output characteristics of various adjustable and controllable resources in active distribution networks are studied.For photovoltaic power generation,the basic principle,influencing factors and modeling method are analyzed.For energy storage system,the specific classification and electrochemical energy storage model are introduced.Meanwhile,the model of micro turbine and the electricity price compensation model of Interruptible load are studied.Finally,the influences of adjustable and controllable resources on power flow,voltage,and network loss in the active distribution network are analyzed in detail by simulation.(2)Robust optimal scheduling in active distribution networks considering temporal and spatial correlation are proposed.Firstly,the necessity of temporal and spatial correlation is verified based on the output data of two actual photovoltaic power stations.The temporal and spatial correlation constraints are modeled using Pearson correlation coefficients,which is nonlinear.The nonlinear constraints can be transformed into linear constraints due to the discrete characteristics of the uncertain set.Next,a robust optimal dispatching model in the active distribution network considering temporal and spatial correlation is established.The model is essentially three-level two-stage “min-max-min” problem,which is hard to solve.The duality theory and lager M algorithm are employed to transform the nonlinear “min-max-min” problem into linear “min-max” problem.Finally,the column constraint generation method(C&CG)algorithm is used to solve the model,and the effectiveness and correctness are verified in an improved IEEE33 system.And the results focus on analyzing the impact of temporal and spatial correlation.(3)The distributed optimal scheduling of active distribution networks is proposed.Firstly,the uncertainty model of photovoltaic are analyzed and the correlation of photovoltaic and load is simulated based on Latin Hypercube Sampling(LHS)and Cholesky decomposition technology.Secondly,the distributed optimal scheduling is established which includes the upper-layer distribution network optimization model and the lower-level microgrid optimization model.Obviously,there are a coupling relationship between the upper-layer distribution network and the lower-level microgrid through the interaction power.Next,the method of copying nodes which connection line information are simultaneously calculated in the distribution network and microgrid model,is used to realize the decoupling of the distribution network and microgrid.Finally,the proposed model is solved by using the analytical target cascading(ATC),and the correctness of the model is verified in an improved IEEE 123 system.The convergence process of the contact line power are emphasized.