20 October 2023 in IEEE Transactions on Circuits and Systems II: Express Briefs
To enhance the robustness and reduce the steady-state error under impulsive noise environment, this paper proposes an augmented complex generalized modified Blake-Zisserman (ACGMBZ) algorithm to estimate the fundamental frequency of three-phase power system adaptively. First, based on Clarke’s αβ transformation, the widely linear (WL) model is introduced to form a complex three-phase voltage for capturing more information than strictly linear (SL) model in unbalanced conditions. Then, by employing the generalized modified Blake-Zisserman (GMBZ) cost function, the ACGMBZ algorithm is derived using CR calculus which provides a mathematical framework for computing the Jacobians, Hessians, and gradients of cost functions of complex variables Mandic2009.. In addition, the steady-state analysis of the proposed algorithm is presented, as well as convergence analysis in both mean and mean-squared senses. Finally, some typical operations of the three-phase power system are considered as numerical simulation examples to demonstrate the improved performance of the proposed ACGMBZ algorithm by both synthetic and real grid data.