Data-Driven Method for Missing Harmonic Data Completion

In large-scale applications, missing harmonic data during transmission is inevitable. This paper presents a novel approach for the completion of missing harmonic data based on a data-driven approach. The idea is to construct an adjacency matrix and unconstrained optimization function through spectral graph theory, and construct an a priori information model of the target bus using the merging K-means algorithm. The augmented Lagrange iteration algorithm is used to iteratively solve the unconstrained optimization function, so to complete the missing harmonic data. In comparison with existing methods, this method reduces the data dependence between different phase measurement units and does not require the optimization strategy of phasor measurement units, which is suitable for single-phasor harmonic data completion. Finally, the effectiveness and practicability of the proposed method were verified using artificial and field data.