Spatial-temporal associations representation and application for process monitoring using graph convolution neural network

Hao Ren,Xiaojun Liang,Chunhua Yang,Zhiwen Chen,Weihua Gui

2023-10-05

Abstract:Thank you very much for the attention and concern of colleagues and scholars in this work. With the comments and guidance of experts, editors, and reviewers, this work has been accepted for publishing in the journal "Process Safety and Environmental Protection". The theme of this paper relies on the Spatial-temporal associations of numerous variables in the same industrial processes, which refers to numerous variables obtained in dynamic industrial processes with Spatial-temporal correlation characteristics, i.e., these variables are not only highly correlated in time but also interrelated in space. To handle this problem, three key issues need to be well addressed: variable characteristics modeling and representation, graph network construction (temporal information), and graph characteristics perception. The first issue is implemented by assuming the data follows one improved Gaussian distribution, while the graph network can be defined by the monitoring variables and their edges which are calculated by their characteristics in time. Finally, these networks corresponding to process states at different times are fed into a graph convolutional neural network to implement graph classification to achieve process monitoring. A benchmark experiment (Tennessee Eastman chemical process) and one application study (cobalt purification from zinc solution) are employed to demonstrate the feasibility and applicability of this paper.

Machine Learning,Artificial Intelligence,Systems and Control

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the spatio - temporal correlation representation of dynamic, correlated and multi - scale variables in modern industrial processes and its application in process monitoring. Specifically, the paper focuses on how to effectively capture and utilize the spatio - temporal correlation knowledge implicit in these variables to achieve efficient monitoring of industrial processes.

### Problem Background

1. **Spatio - temporal Correlation Characteristics**:
- A large amount of data generated in modern industrial processes is dynamic and correlated. These data are highly correlated not only in time but also in space.
- When an anomaly occurs, the data distribution of the same variable in different time periods will change irregularly, and the relationships between different variables will also change.

2. **Challenges**:
- Data knowledge in industrial processes is usually closely related to mechanism and process knowledge and contains richer semantic information.
- Any simple anomaly or error may lead to serious consequences, so the reliability requirements for data knowledge are extremely high, making data mining extremely difficult.

### Solutions

To address the above problems, the paper proposes the following solutions:

1. **Variable Feature Modeling and Representation**:
- Assume that the data follows an improved Gaussian distribution to model variable features.

2. **Graph Network Construction (Time Information)**:
- Use monitoring variables and their time features to calculate edges and construct static graph network snapshots.

3. **Graph Feature Perception**:
- Input the process state graph networks corresponding to different time points into the Graph Convolutional Neural Network (GCN) to achieve graph classification, thereby achieving the purpose of process monitoring.

### Application Verification

The paper verifies the effectiveness and applicability of the proposed method through two case studies:

1. **Benchmark Experiment**:
- Use the Tennessee Eastman chemical process as a benchmark. This process involves 52 monitoring variables, simulates 21 kinds of faults and one normal state to verify the feasibility and effectiveness of fault detection and diagnosis.

2. **Practical Application Research**:
- In the zinc hydrometallurgical process, cobalt ions are removed by adding impurity - removing agents and catalysts. 49 variables are used for long - term stable operation control, and 8 operating conditions are defined to verify the effectiveness and feasibility of the proposed method in the actual industrial process.

### Conclusion

The method based on spatio - temporal correlation representation proposed in the paper can effectively capture the irregular dynamic changes between variables, achieve fault diagnosis accuracy comparable to or even better than existing methods, and demonstrate its application potential in actual industrial processes.