Abstract: To ensure the safety of pressure pipeline, inspectors and users need to carry out inspection and maintenance according to relevant laws and standards. However, due to the huge and complex legal system, it is extremely necessary to develop a knowledge traceability system in the field of pressure pipeline inspection. Combined with the pressure pipeline inspection knowledge system, inspection knowledge data was intelligently updated by using data crawler and OCR technology. Because of the ability of knowledge graph to efficiently integrate massive knowledge information, a pressure pipeline inspection knowledge graph construction technology was proposed to construct triple and complete knowledge extraction based on Bert-BiLSTM-CRF knowledge extraction method. The map construction and visualization of pressure pipeline inspection knowledge were completed on Neo4j platform. A traceability system for pressure pipeline inspection knowledge atlas was designed and developed, and the accuracy and practicability of the traceability system were verified through retrieval and application tests.