Abstract
By combining together the extended Kalman filter with a newly developed C&I particle swarm optimization algorithm (C&I-PSO), a novel estimation method is proposed for parameter estimation of electromechanical oscillation, in which critical physical constraints on the parameters are taken into account. Based on the extended Kalman filtering algorithm, the constrained parameter estimation problem is formulated via the projection method. Then, by utilizing the penalty function method, the obtained constrained optimization problem could be converted into an equivalent unconstrained optimization problem; finally, the C&I-PSO algorithm is developed to address the unconstrained optimization problem. Therefore, the parameters of electromechanical oscillation with physical constraints can be successfully estimated and better performed. Finally, the effectiveness of the obtained results has been illustrated by several test systems.
Highlights
Over the past few decades, there have been a lot of concerns about estimation of electromechanical modes since it can offer a substantial amount of important information about power system stability [1,2,3,4,5]
One of them is the ambient data, which can be sampled from a power system at a stable operating point [10]; the other one is designed as the ringdown data, which is generated from a power system with a major disturbance
In order to evaluate the whole performance of the proposed approach and conventional EKF method effectively, the root-mean-square deviation (RMSD) used in [43] is adopted: RMSD =
Summary
Over the past few decades, there have been a lot of concerns about estimation of electromechanical modes since it can offer a substantial amount of important information about power system stability [1,2,3,4,5]. A novel method for ringdown detection was proposed based on the traditional EKF in [3], where the damping ratio and frequency of ringdown data were modeled in the state vector; these parameters could be estimated by using EKF. In this paper, a modified PSO algorithm (namely, C&I-PSO) is developed to address the equivalent unconstrained optimization problem, in which a constrictive factor and a linear decreasing inertia weight (LDIW) are introduced to improve the performance of global searches and local searches [42] By this method, in the early search stage, a large inertia weight is utilized to extend the search region and avoid the occurrence of premature problems.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call