Information Reconstruction Error Evaluation Method Based on Singular Entropy Method

Communication failures of digital substations can result in loss of sampled information data. Information reconstruction algorithms are often used to reconstruct missing sample data, which causes errors in the amplitude and phase of the Fourier filter calculation. Studies have shown that the amount and location of sampled data loss has a direct impact on the size of the reconstruction error. Combining the singular value decomposition and information entropy theory, this paper comprehensively considers the influencing factors of reconstruction error, studies the correlation between reconstruction error and sampling data singular entropy, and proposes a reconstruction error estimation method based on singular entropy. This method is used to evaluate the reconstruction errors of three interpolation methods.