A Variable Structure Trajectory Predictive Algorithm Based on Transform For Complex Exponential Times Series

With the application of APMU (angle & phase measurement unit) based on GPS (global position system) in power system, online system trajectory prediction that is of great significance to transient stability emergent control becomes feasible. Based on normal form (NF) analysis at the instant of removal of the fault, a predictive algorithm for complex exponential time series is proposed in this paper, then considering trajectory prediction characteristic and the difference of using observation data length, a variable structure predictive algorithm is advised. The proposed prediction method is based on Prony´s nonlinear approximating. Prony method and their recent extensions display the feasibility for online power system trajectory approximating. The trajectory prediction procedure and results are illustrated using computer simulation of 36 buses system with 8 generators (EPRI-36 China). Numerical simulation results have shown that the proposed method is effective and practically feasible