Power distribution network dynamic state estimation method and system based on square root cubature Kalman filter

Abstract

The invention discloses a power distribution network dynamic state estimation method and system based on a square root cubature Kalman filter. The power distribution network dynamic state estimation method comprises the following steps: S1, acquiring measurement data at a certain moment k; S2, performing state prediction according to the measurement data at the moment k, and predicting the state at the moment k + 1 by using a holt two-parameter exponential smoothing method to obtain a state prediction value at the moment k + 1; S3, correcting the state prediction value at the k + 1 moment by using the measurement data to obtain a state estimation value at the k + 1 moment; S4, updating filtering parameters to prepare for next iteration; carrying out the next iteration, returning to the step S1, obtaining measurement data at the moment k + 1, and executing the steps S2-S4 to realize state estimation at the moment k + 2. According to the invention, dynamic state estimation can be carried out by using the collected power distribution network measurement data to obtain more accurate operation state information of the power distribution network, and measurement errors caused by measurement equipment are reduced. Compared with a standard volume Kalman filtering dynamic state estimation algorithm, the algorithm is stronger in robustness and higher in state estimation precision.

Description

technical field [0001] The invention relates to distribution network operation analysis and management, and distribution network dynamic state monitoring, in particular to a distribution network dynamic state estimation method and system based on a square root volumetric Kalman filter. Background technique [0002] The state estimation of the distribution network is to use the measurement data collected by the data acquisition system to estimate the optimal state of the system, which is of great significance to the analysis and management of the distribution network operation. The traditional power system state estimation is considered as a static estimation problem, which is solved by weighted least squares method in most cases. With the deployment of phasor measurement units, the update frequency of measurement data is greatly increased, and dynamic state detection of power system status has become possible. Traditional static state estimation methods can only estimate th...

AI Technical Summary

Problems solved by technology

[0003] Existing problems: Generally speaking, in the traditional unscented Kalman filter sigma sampling model, sampling is performed assuming that the parameter values ​​​​are fixed

However, the algorithm based on volumetric Kalman filter will appear asymmetric or non-positive definite covariance matrix phenomenon during the algorithm iteration process, which will reduce the estimation accuracy and even interrupt the iteration process.

Images