Data-driven prediction of heat exchanger steady-state conditions based on support vector regression

Determining the steady-state operating conditions of heat exchangers is both a fundamental and complex computational task. Traditional methods typically depend on intricate theoretical models and extensive empirical data, making them labor-intensive and less efficient. Conversely, machine learning and data-driven methods exhibit advantages in simplicity, ease of use, and rapid response capabilities. This research explores the application of Support Vector Regression (SVR) enhanced by Grey Relational Analysis (GRA) to predict steady-state conditions of preheaters and steam generators. The results reveal that SVR is remarkably effective in forecasting the steady-state operations. Although predictive performance varies between the preheater and steam generator, the proposed model consistently achieves high accuracy, with mean errors below 3.5 %. Furthermore, the model exhibits strong robustness, maintaining reliability despite input fluctuations, thereby highlighting its potential for practical deployment in real-world settings.