Spectral Variation-Based Compression: A Case Study with Leakage Current Monitoring

This work presents a spectral variation-based compression method for leakage current signals from high-voltage insulators. This scheme is applied to real data acquired from sensors installed at Hub Sergipe. We propose the Spectral Variation Detection Algorithm (SVDA), adapted from the Spectral Variation Detection and Compression System (SVDCS) and implemented on a resource-constrained IoT device. The algorithm optimizes storage by leveraging the Hermitian symmetry property, enabling efficient frequency-domain compression. Using signals digitized at 7.68 kHz, the best compression ratio achieved with the default threshold was 38.7:1. This method provides a practical solution for compressing leakage current signals in real-world applications.