Research On Dynamic Order Reduction Method And Applicability For Wind Power Grid Connected System Under Large Disturbance

With the proposal of the "30·60 dual carbon" goal,China’s wind power industry has flourished,and the installed scale,power generation,and utilization rate of wind power have repeatedly reached new highs.As a new energy generation technology,wind power generation is constantly increasing its grid connection scale,but it also poses serious challenges to the safe and stable operation of the power system.The high penetration rate of wind power integration has led to profound changes in the transient characteristics of new energy grids,endangering the safe and stable operation of the power system.Therefore,it is urgent to carry out research on the transient characteristics of wind power grid connected systems.However,if every wind turbine unit is modeled with detailed models,it will lead to many problems such as curse of dimensionality,time scale span increase,etc.Therefore,establishing a reduced order model of wind power suitable for large disturbance analysis is the basis and key to carry out the research on the transient characteristics of wind power grid connected systems.Therefore,this article takes the direct drive permanent magnet synchronous generator(PMSG)wind power generation system as the research object,and studies the dynamic reduction method of the direct drive wind power grid connected system suitable for large disturbance analysis.The following aspects of research are carried out,specifically as follows:(1)A Singular perturbation order reduction method for PMSG unit level based on single point linearization model is proposed.Firstly,the basic structure of PMSG and the dynamic mathematical model of key components were presented,providing a research foundation for subsequent model reduction of PMSG.Then,according to the multi time scale characteristics of the wind power grid connected system,a PMSG Singular perturbation order reduction method based on the linearization model of the balance point is proposed.The participation factor analysis method is used to calculate the state variables strongly related to each dominant mode.According to the degree of correlation between the dominant mode and each state variable,fast and slow variables are divided.Finally,the PMSG Singular perturbation order reduction model is established,and the effectiveness of the proposed order reduction method is verified by a simulation example.(2)On the basis of trajectory piecewise linearization technology,in order to address the problem of existing model reduction methods being difficult to accurately characterize the strong nonlinear characteristics of wind power grid connected systems in large disturbance scenarios,a PMSG nonlinear dynamic reduction method based on disturbance trajectory space decoupling is proposed through a multi-point linearization model.Firstly,based on the trajectory piecewise linearization technique,a local linearized model of PMSG was obtained through the perturbed trajectory of PMSG.Each local linearized model was spatially decoupled,and a generalized participation matrix for the dominant eigenvalues of the local linearized model and the full order system state variables was proposed.Then,according to the generalized participation factor,the PMSG system is divided into multiple time scales,and then the PMSG Singular perturbation reduced order model suitable for large disturbance scenarios is established.Finally,simulation has demonstrated that the proposed reduced order model can ensure high accuracy,computational efficiency,and applicability even when the testing system is subjected to large disturbances.(3)In order to solve the problems of curse of dimensionality and low simulation efficiency if every unit in the wind farm is modeled with detailed models,a model reduction method based on multi trajectory piecewise linearity is proposed to reduce the order of direct driven wind farms under large disturbances.Firstly,the basic idea of the multi trajectory segmented linear model reduction method was introduced,and the PMSG multi trajectory segmented linear model was derived.Then,the key technologies for implementing the multi trajectory segmented linear model reduction method were introduced in detail.Finally,the accuracy and robustness of the proposed direct drive wind farm MTPWL model in capturing the transient behavior of PMSG wind farm nonlinear models under large disturbances were verified by applying different large disturbances to the testing system.