Pretrained Language–Knowledge Graph Model Benefits Both Knowledge Graph Completion and Industrial Tasks: Taking the Blast Furnace Ironmaking Process as an Example

Xiaoke Huang,Chunjie Yang

DOI: https://doi.org/10.3390/electronics13050845

IF: 2.9

2024-02-23

Electronics

Abstract:Industrial knowledge graphs (IKGs) have received widespread attention from researchers in recent years; they are intuitive to humans and can be understood and processed by machines. However, how to update the entity triples in the graph based on the continuous production data to cover as much knowledge as possible, while applying a KG to meet the needs of different industrial tasks, are two difficulties. This paper proposes a two-stage model construction strategy to benefit both knowledge graph completion and industrial tasks. Firstly, this paper summarizes the specific forms of multi-source data in industry and provides processing methods for each type of data. The core is to vectorize the data and align it conceptually, thereby achieving the fusion modeling of multi-source data. Secondly, this paper defines two interrelated subtasks to construct a pretrained language–knowledge graph model based on multi-task learning. At the same time, considering the dynamic characteristics of the production process, a dynamic expert network structure is adopted for different tasks combined with the pretrained model. In the knowledge completion task, the proposed model achieved an accuracy of 91.25%, while in the self-healing control task of a blast furnace, the proposed model reduced the incorrect actions rate to 0 and completed self-healing control for low stockline fault in 278 min. The proposed framework has achieved satisfactory results in experiments, which verifies the effectiveness of introducing knowledge into industry.

engineering, electrical & electronic,computer science, information systems,physics, applied

What problem does this paper attempt to address?

The main aim of this paper is to address the application challenges of Industrial Knowledge Graphs (IKGs) in industrial tasks, particularly focusing on the Blast Furnace Ironmaking Process (BFIP). Specifically, the research focuses on the following key issues:

1. **Multi-source Data Integration**: The paper proposes methods for handling different forms of industrial data, including structured data, semi-structured data (such as local knowledge graphs), and unstructured data (such as production logs and expert knowledge documents). The goal is to achieve effective integration of these data by converting them into vector representations.

2. **Building High-Quality Industrial Knowledge Graphs**: The research introduces a pre-trained language-knowledge graph model that aligns entity nodes with textual concepts. By utilizing entity prediction and link prediction tasks, the model extracts and integrates information from textual and graphical data, thereby improving the quality of the knowledge graph.

3. **Two-Stage Model Construction Strategy**: First, production logs and static IKGs are used as a knowledge base to build a pre-trained model. Second, based on the pre-trained model, the model is further optimized through a multi-task learning framework by incorporating sensor data required for dynamic industrial tasks and dynamic subgraphs related to specific tasks, to adapt to different industrial application scenarios.

4. **Enhanced Dynamic Adaptability**: To improve the model's adaptability to dynamic industrial tasks, the research introduces a dynamic expert network structure for fine-tuning. This helps the model better handle dynamic data in industrial processes and provides more precise solutions for specific tasks (such as fault diagnosis, self-healing control, etc.).

In summary, the main objective of this paper is to overcome the challenges in the application of existing industrial knowledge graphs by constructing a model that can effectively handle multi-source data, complete knowledge graph supplementation, and be applied to various industrial tasks.