Power system dynamic state estimation method based on singular value decomposition CDKF

Abstract

The invention discloses a power system dynamic state estimation method based on singular value decomposition CDKF. First, a singular value decomposition (SVD) technology is introduced to central difference Kalman filtering CDKF, and Collins basic decomposition of a covariance matrix in CDKF is replaced with singular value decomposition. Then, with the help of the Stering interpolation formula, the derivation of a nonlinear function is approximated by a polynomial, and a first-order or second-order derivation formula in a Taylor expansion is replaced with a central difference formula. Through the method, the problem that the covariance matrix is not positive definite due to calculation error and rounding error is solved. The numerical stability of the method is improved. Complex deviation is avoided. Calculation is simple, parameter adjustment is easy, and the filtering precision is high. The results show that the method has higher numerical stability than a square root form method and has higher filtering precision than an extended Kalman filter method and an unscented Kalman filter method.

Description

technical field [0001] The invention relates to a method for estimating the dynamic state of a power system based on singular value decomposition CDKF, and belongs to the technical field of power system monitoring, analysis and control. Background technique [0002] With the continuous expansion of the scale of the power system, the complexity of the power grid continues to increase. Because the power system operating state data obtained by direct measurement means contains measurement errors, it cannot be directly used as data support for monitoring and analysis. Power system dynamic state estimation can filter out measurement errors in direct measurement data, and has the ability to predict the operating state of the power system at the next moment, and can obtain more accurate state information, which is indispensable in the analysis of power system operation. status. [0003] An important link in dynamic state estimation is to establish a stable filtering method with h...

AI Technical Summary

Problems solved by technology

The extended Kalman filter performs the first-order truncation of the Taylor expansion of the nonlinear function, and needs to calculate the Jacobian matrix. The filtering accuracy is low and the calculation amount is large; although the unscented Kalman filter does not directly approximate the nonlinear equation, the filtering accuracy is higher than that of the extended Kalman filter. Kalman filter

However, in the process of generating sampling points in the unscented Kalman filter, the covariance matrix is ​​decomposed by Kolinsky. Due to the influence of calculation errors and rounding errors, the covariance matrix is ​​easy to lose positive definiteness during the calculation process, resulting in interruption of the filter, or Filter divergence due to inaccurate process noise covariance

Images