Wind Farm And Branch State Estimation Based On Cubature Kalman Filter

Wind power grid-connected operation is an effective way for large-scale development and utilization of wind energy.Wind power is characterized by volatility and incomplete control.With the incorporation of a large number of wind farms into the power grid,wind farms have an increasingly greater influence on the stability of the entire power network.How to quickly find the fault point and quickly remove the fault in the transient process of the failure of the wind farm becomes a problem.At the same time,the active and reactive dispatches of the wind farm are dependent on the effect of state estimation.Therefore,this paper has a clear academic significance and engineering value based on the state estimation of the wind farm.The traditional research on state estimation of wind farms focuses on the state estimation of grid point and emphasis on the state estimation from the perspective of the grid system.This paper starts from the characteristics of the internal structure of the wind farm and conducts research on the state estimation of each node within the wind farm.Its main work is reflected in the following aspects:(1)Starting from the wind farm network structure,the dynamic characteristics of each node of the system are analyzed,and the equivalent modeling of the wind farm is performed based on the typical network structure.(2)In order to estimate the state of grid-connected wind farms,combined with Holt's two-parameter exponential smoothing method and the equivalent model of wind farm,a system state space model was established and a square root kalman filter was designed.Through numerical analysis: Compared with the existing methods,the posterior mean and covariance are significantly reduced,which improves the accuracy of state estimation.(3)Aiming at the characteristics that the wind speed influence the wind power branch fluctuate greatly,a strong tracking square root cubature kalman filter is designed to improve the precision of state estimation;and further,a strong tracking based on the distributed structure of wind farm is proposed square root kalman filter for distributed cooperative parallel state estimation algorithm and distributed cooperative sequence state estimation algorithm.The numerical experiments show that in the uncertain environment of wind farms,the cubature kalman filter can estimate the effect better,thus verifying the effectiveness of the algorithm.