Research on intelligent analysis and prediction of low-voltage causes in rural distribution networks based on deep learning

Abstract The article introduces an advanced diagnostic approach for identifying the causes of low voltage in power distribution networks. This method integrates empirical analysis of low-voltage causes and employs both the particle swarm optimization-enhanced support vector machine algorithm and the self-organizing map algorithm to forecast low-voltage events within the distribution network. By merging low-voltage cause analysis with machine learning algorithms, it achieves precise diagnostics of low-voltage issues. This methodology has demonstrated remarkable performance in real-world applications across multiple regions, effectively pioneering an automated diagnostic technology for detecting low-voltage problems in power distribution networks.