Power System Dynamic State Estimation Based On Uncertainty Measurement

The construction of smart grid requires the dispatching center to quickly and accurately sense the state of the grid and judge the development trend.The power system state estimation is the key technology to achieve the above requirements.Static estimation represented by weighted least squares is widely used in power systems,but such methods have long calculation periods and high requirements for measurement data integrity.The dynamic estimation method can overcome the above shortcomings and predict the state change of the grid at the next moment,so it is more in line with the real-time requirements of the smart grid.However,power system dynamic estimation has problems such as dynamic model simplification,measurement data heterogeneity,and bad data identification in practical applications.In order to solve the above problems,this paper studies the static estimation algorithm,dynamic estimation model and dynamic estimation algorithm,and proposes a dynamic estimation method based on measurement uncertainty.In the aspect of static estimation method,the maximum static estimation method of the normal rate of the measurement point based on the measurement uncertainty is introduced,and the statistical interpretation of the maximum likelihood method with the normal rate of the measurement point as the random variable is given.The differences between measurement error,estimation error,residual error and measurement uncertainty are analyzed.Then a evaluation function in the form of cosine function is proposed.In the aspect of dynamic estimation model,a recursive extended parameter model based on fading memory method is proposed.The state transition function is simplified by diagonal matrix and the recursive least squares method is used to identify and update online.The dynamics of forgetting factor on grid state is introduced.The change is tracked;the time-variable measurement model in the case of heterogeneous measurement data is proposed,and the corresponding measurement equation is established according to the type and quantity of the measurement data at the current time,and the dimensional consistency of the matrix operation due to the change of the matrix dimension is discussed.The formula is derived based on the maximum a posteriori estimation criterion.In the aspect of dynamic estimation algorithm,the above model is combined with the unscented Kalman filter method to construct the dynamic estimation framework.According to the principle of unscented transformation,the sampling strategy suitable for the power system is selected.Finally,the dynamic estimation update step with strong robustness is proposed.The information vector is abnormal,and the bad data is identified by the static estimation based on the measurement uncertainty.Then the update factor matrix is used to correct the filter gain,and the influence of the bad data in the updating step is suppressed.The simulation example shows that the proposed method has high estimation accuracy and good robustness.