Digital Transformation in Power Transformer O&M: Critical Technologies for Multi-Source Heterogeneous Data Lifecycle Processing

Multi-source heterogeneous data processing for power transformers serves as a critical driver for the digital transformation of operation and maintenance (O&M). This paper systematically elaborates key technologies for full lifecycle data processing. Firstly, it defines the characteristics of multi-source heterogeneous data in power transformers. Subsequently, preprocessing methods are proposed, including text and audio data cleansing, as well as structured data normalization. Furthermore, an analytical framework is established that integrates structured data analysis and text mining techniques. The paper discusses data fusion technologies and demonstrates the effectiveness. Finally, future trends are prospected, highlighting the application of big data processing and large language models as pivotal research directions for advancing intelligent lifecycle management.