Reactive Power Optimization Of Wind Farm In System Restoration Based On Model Predictive Control

With the continuous expansion of power grid scale,the operating environment of power system is also getting increasingly complex,all kinds of sudden blackout accidents of power system can't be completely avoided.During the restoration of the blackout system,the capacity of the no-load line and a large number of inductive load input will contribute the impact on the reactive power distribution of the restored system,thus cause the system node voltage deviation or even violate the limit,which has an adverse impact on the security restoration of power system.At present,the reactive power flow of power system is often optimized by adjusting the reactive power output of conventional generator sets in the process of system restoration,but the reactive power regulation capacity of conventional generator is limited and the transmission of long distance reactive power will cause a lot of line active power loss.With a large amount of wind power in the power system integrated to the grid through power electronic converter,which renders the wind farms has independent active and reactive power control ability and be able to quickly provide reactive power support for the power system to ensure the voltage stability of the restored system.For this purpose,this paper came up with the dispatch method for wind farm reactive power based on model predictive control,and the research on the reactive power optimization strategy for wind farm participating in system restoration,this article has mainly completed work as followings:Firstly,it analyzed both traditional reactive voltage regulating equipment in power system operation and the characteristics of new types of reactive power regulating equipment controlled by power electronic device,the principle of reactive power regulation and its application fields are summarized,in which focused on analyzing the reactive power output characteristics of conventional wind turbines in the current stage,and the reactive power output models of these wind turbines are derived and calculated.Secondly,it studied the real-time reactive power output optimization model for wind farm participating in system restoration with the objective of minimizing the weighted sum value of the system active power loss and node voltage deviation,first,regarded the doubly-fed inductive generator as an example,the reactive capacity of wind farm influenced by active power output changes of wind farm has been analyzed.The range of reactive power output of the whole wind farm has been calculated according to the wind turbine output model.The doubly-fed inductive generator wind farm has been deemed as a continuous reactive power source that participating in the reactive power optimization during system restoration,and then transformed real-time reactive power optimization problem about the doubly-fed inductive generator wind farm participating in the system restoration process into the nonlinear optimization problem with multiple constraints and solve the problem by using Artificial Bee Colony optimization algorithm,Finally,the effectiveness of the wind power participating in real-time reactive power output optimization model for system restoration is verified by IEEE-39 nodes system.Thirdly,in consideration of when the wind farm was involved in the system restoration process,the real-time reactive power optimization method can not adapt to the disturbance and error caused by system state change in the process of power system restoration,the model predictive control(MPC)method is introduced and a multi-step rolling optimization model of reactive power output based on MPC was established as well as proposed the reactive power optimization control strategy for restored system under different voltage excursions,the multi-step rolling optimization control was carried out for the variable quantity of reactive power output regulation of wind farm at each time step according to the new method for calculating the sensitivity of voltage or system active power loss to reactive power increment,the quadratic programming software was used to solve the problem and regarded it's result as the reference value of reactive power adjustment for each time step and the system active power loss and voltage variation of the continuous time step of the system are obtained,in the end,the IEEE-39 nodes system and the actual system of a regional power system have been saw as the example validates based on the effectiveness of the reactive power output optimization model based on MPC with wind power participating in the system restoration process,and compared it with the real-time reactive power optimization results of the existing wind farm participating in system restoration,which are proved that the wind farms participating in reactive power optimization control method in system dynamic restoration based on the MPC is better.