An Efficient Prediction Method for Boiler Water Wall Temperature Based on Thermal Inertia Modeling and State-Space Enhancement
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
To improve the prediction accuracy and engineering practicality of boiler water wall temperature in power plants, this paper proposes a state-space-enhanced prediction model that integrates thermodynamic mechanisms with first-order lag characteristics. Starting from the fundamental laws of heat transfer, a discrete state-space expression for the dynamic response of wall temperature is derived, theoretically elucidating the coupling mechanism between the influence of operating parameters and the thermal inertia effect of the metal wall. Based on this theory, a “minimal feature enhancement” prediction method is constructed: within a multiple linear regression framework, only the wall temperature from the previous moment is introduced as an additional feature, extending the model to a first-order autoregressive form. Validation using actual operational data from a$\mathbf{6 4 0 ~ M W}$supercritical unit shows that the proposed state-space-enhanced model achieves a prediction root mean square error of 0.22° C and a coefficient of determination of 0.9990. Compared to a baseline model using only operating parameters, the error is reduced by approximately 70%, while computational complexity increases only marginally. By embedding simple and interpretable features, this method effectively characterizes the system's thermal inertia, providing a high-precision, easily deployable solution for boiler safety monitoring.
Fields of Study & Subjects
Advanced Power Generation Technologies
Fault Detection and Control Systems
Advanced Control Systems Optimization
Publication details
- Journal / source
- International Conference on Electrical Automation and Artificial Intelligence (ICEAAI)
- Pages
- 128–132
- Publication date
- 16 January 2026
- Language
- EN