Error Estimation Method of Reduced-Order Small-Signal Model for Multiterminal DC Distribution Network

To ensure the accuracy of the application using the reduced-order small-signal model in multiterminal dc distribution network dynamic analysis, an efficient error estimation method for the eigenvalues that represent the dominant modes is proposed in this article based on the matrix perturbation theory and Newton algorithm. To further improve the estimation efficiency and provide a basic judgment of error trend with respect to the change in parameters, an error bound estimation method and a sensitivity expression of reduction error are developed, respectively. The verification results on two typical linear time-invariant systems and a practical three-terminal dc distribution system demonstrate that the proposed estimation method is of high accuracy and efficiency in quantifying the error/bounds of the dominant modes, and the derived sensitivity indicates the trend of eigenvalue error concerning parameters and the rationality of order reductions.