Open Access

ARTICLE

Construction Method of Equipment Defect Knowledge Graph in IoT

Huafei Yang¹, Wenqing Yang¹, Nan Zhang¹, Shanming Wei^2,*, Yingnan Shang¹

1 NARI Group Corporation (State Grid Electric Power Research Institute), Nanjing, 211106, China
2 School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, 210094, China

* Corresponding Author: Shanming Wei. Email:

Intelligent Automation & Soft Computing 2023, 37(3), 2745-2765. https://doi.org/10.32604/iasc.2023.036614

Received 06 October 2022; Accepted 13 December 2022; Issue published 11 September 2023

Abstract

Equipment defect detection is essential to the security and stability of power grid networking operations. Besides the status of the power grid itself, environmental information is also necessary for equipment defect detection. At the same time, different types of intelligent sensors can monitor environmental information, such as temperature, humidity, dust, etc. Therefore, we apply the Internet of Things (IoT) technology to monitor the related environment and pervasive interconnections to diverse physical objects. However, the data related to device defects in the existing Internet of Things are complex and lack uniform association hence building a knowledge graph is proposed to solve the problems. Intelligent equipment defect domain ontology is the semantic basis for constructing a defect knowledge graph, which can be used to organize, share, and analyze equipment defect-related knowledge. At present, there are a lot of relevant data in the field of intelligent equipment defects. These equipment defect data often focus on a single aspect of the defect field. It is difficult to integrate the database with various types of equipment defect information. This paper combines the characteristics of existing data sources to build a general intelligent equipment defect domain ontology. Based on ontology, this paper proposed the BERT-BiLSTM-Att-CRF model to recognize the entities. This method solves the problem of diverse entity names and insufficient feature information extraction in the field of equipment defect field. The final experiment proves that this model is superior to other models in precision, recall, and F1 value. This research can break the barrier of multi-source heterogeneous knowledge, build an efficient storage engine for multimodal data, and empower the safety of Industrial applications, data, and platforms in multi-clouds for Internet of Things.

---

Keywords

---