Voltage Optimal Control Of AC/DC Hybrid Active Power Distribution Network Based On Model Predictive Control

In order to meet the needs of the society for clean energy,the distribution network must accept a large number of distributed generators in the future.AC/DC hybrid active power distribution network topology flexible,power flow control,which can accept more distributed generators.AC/DC hybrid active power distribution network will become an important part of the future distribution network.Compared with the transmission network,the reactance is no longer much greater than that of resistance in distribution network,so the change of the active power output of the distributed generators will affect the amplitude of the voltage.The traditional voltage optimal control method is based on the accurate prediction of the load and the distributed generators.But the distributed generator output has the characteristics of volatility and intermittent,which is difficult to predict,low accuracy,and the accuracy of load estimation in distribution network is also poor.After the high penetration of distributed generators,the uncertainty of system operation is more serious.The traditional voltage control method of distribution network has been challenged.It is urgent to come up with a control method of voltage optimization to cope with the strong uncertainty in the distribution network.Based on the consideration of the uncertainty of distributed generators and load,this paper proposes a method of voltage optimization control for hybrid active power distribution network based on model predictive control.The AC/DC hybrid active power distribution network this paper studied is composed by flexible multi terminal DC device.The mathematical model of flexible multi terminal DC device is deduced based on the mathematical model and control method of VSC.And studied how the model predictive control theory is combined with the voltage control.On the basis of these theoretical studies,a mathematical model for the optimal control of AC/DC hybrid active power distribution network based on model predictive control is built.The model consists of three parts,including prediction model,rolling optimization and feedback correction.The variation of distributed generator and load is different,so this paper uses different forecasting methods.The rolling optimization model is a nonlinear programming problem which included the discrete variables and continuous variables,and contains active power and reactive power.In this paper,the interior point method and the nonlinear complementarity constraints are used to solve the problem.Depending on whether the node voltage is off limited,the feedback correction is selected to correct the control command or only modify the prediction model.In addition,in order to speed up the convergence,some constraint conditions are relaxed in order to speed up the convergence.In order to avoid the frequent movement of the discrete equipment,the feedback correction is a priority to adjust the continuous reactive power equipment.In order to verify the correctness and effectiveness of the proposed method,the IEEE standard network and the actual power grid of Yanqing are used to verify the correctness and effectiveness of the proposed method.And compared with the traditional control methods.The simulation results show that,the control method proposed in this paper can deal with the uncertainty of the distributed generators and load,and keep the voltage of each node of the whole network in a reasonable range.In addition,it can effectively reduce the active power loss of the whole network.