The invention belongs to the technical field of power
system monitoring and fault diagnosis, and particularly relates to a power fault
signal denoising method and equipment. The method comprises the following steps: firstly segmenting a noisy
signal, and mapping each sampling point into a superposed
quantum state representing two possibilities of a fault and
noise; secondly, generating attention distribution by analyzing local uncertainty,
mutation degree and spectral characteristics of the
signal, and guiding intelligent collapse of the
quantum state according to the attention distribution to obtain a preliminary identity mark; then, multi-scale singularity features are extracted through
continuous wavelet transform, and by verifying the physical propagation law of a modulus maximum chain, the preliminary marks are corrected and refined, real fault points are strengthened, and
noise pseudo features are removed; and finally, reconstructing and fusing signals based on the refined state. The method has the adaptive focusing capability of data driving and the reliability of
physical verification, and can extract weak fault transient characteristics in a high-fidelity manner under a strong
noise background.