Research On Optimal Configuration And Operation Strategy Of Distribution Network Containing Shore Power And Hybrid Energy Storage

With the improvement of the electrification level of transportation,the coverage rate of shore power gradually increases,and the impact load brought by shore power gradually increases.Energy storage is connected to the port distribution network to reduce the influence of impact load on power quality.The result of hybrid energy storage capacity configuration affects the power quality improvement effect.The optimal operation results of the distribution network with hybrid energy storage and shore power also affect the power quality improvement effect.Therefore,it is necessary to study energy storage capacity configuration and system optimization operation.However,due to the uncertainty of impact load,it is necessary to study the method of load prediction firstly,to provide data support for optimal operation.To reduce the prediction accuracy due to the uncertainty of impact load,this thesis proposes a combined prediction method.Firstly,the Monte Carlo method is used to simulate the "charging" behavior of ships based on the data of ship inflow and outflow subject to Poisson distribution,to provide ship characteristic data for load prediction.Secondly,an algorithm combined with Wavelet Packet Decomposition(WPD)、Deep Belief Network(DBN)algorithm,and Back Propagation(BP)is used for load prediction to improve the accuracy of load prediction.This thesis proposes a capacity configuration method.Firstly,this method uses the improved ensemble empirical mode decomposition(MEEMD)method to generate several Intrinsic Mode Function(IMFs)and a residual component.Sencondly,this method allocates IMFs to supercapacitors and batteries.Finally,this method solves the nonlinear programming problem by yalmip+fmincon to minimize daily comprehensive costs and with the constraint of the energy storage state of charge(SOC).Aiming at the optimal operation problem of hybrid energy storage connected to the distribution network with shore power,this thesis first establishes the shore power price model,including the dynamic price model and shore power service fee model.Secondly,a two-stage optimization model is established.In the day-ahead optimization stage,the operation cost of the system is minimized to reduce the cost of shore power.In the within-day optimization stage,the hybrid energy storage is dispatched based on the day-ahead optimal operation results,and the load fluctuation is minimized to cope with the change of impact load at a small time scale.Model Predictive Control(MPC)method is used for rolling optimization,which decreases the error of the open-loop optimization method in an uncertain environment.In this thesis,CPLEX + YALMIP is used to solve the mixed-integer programming problem.