A Study of Parallel Configurable Collection Methods for Grid Data Information Containing Imbalance Samples
Desheng Zhou, Liqiang Yao, Zhen Shang, Yiheng Wang, Bo Chen · 2025
The current implementation of data synchronization mining and verification for configurable collection, processing, and sorting is mainly based on single cycles, resulting in low efficiency in data sorting. It is difficult to parallelize configurable collection in power grid data information containing imbalanced samples. Therefore, a parallel configurable collection method for power grid data information containing imbalanced samples is proposed. Using a multi-level approach, complete the multi-level sorting of parallel configurable data information. Based on this, a parallel configurable collection model for unbalanced sample power grid data information is constructed, and data synchronization mining and verification are used to achieve configurable collection and processing. The test results show that the relative error obtained by this design method is relatively small, which can improve the efficiency of data collection in complex background environments.