Research on Intelligent Utilization Technology of Multimodal Archival Data of Power Grid Company Based on Deep Learning of Artificial Intelligence

Grid company archive data has the characteristics of multimodality, mass and diversity, covering structured, semi-structured and unstructured data. These data have the problems of low utilization efficiency and serious information silo effect under the traditional management mode. In order to solve the above problems, this paper proposes a multimodal archive data intelligent utilization technology framework based on artificial intelligence deep learning. Firstly, the convolutional neural network is used to extract features from the image data in the archive to realize image classification and target detection; secondly, the BERT model and the Transformer structure are used to process the archive text data for semantic extraction and text classification; finally, the multimodal Transformer model is used to realize the joint modeling and information sharing of the cross-modal data to solve the problem of isolated and correlation analysis between the data. isolation and correlation analysis. This study provides technical support and application demonstration for the intelligent management and intelligent transformation of power grid company archive data.