Harmonic state estimation based on graph signal processing

This paper presents a novel harmonic state estimation (HSE) methodology based on graph signal processing. The proposed approach can perform HSE without calculating harmonic impedance. The proposed methodology involves graph signal theory for constructing a graph Laplacian matrix and an unconstrained optimization function, the adaptive K-means algorithm for building a prior information model of the target bus, as well as data fuzzification and the domain averaging method for state estimation. Simulations are carried out using a 30-bus power distribution system, and results are presented and discussed to prove the efficiency of the proposed method. The practical application further verifies the availability of the proposed method.