High Pressure Heater Fault Monitoring and Early Warning Based on PCA-XGBoost

In order to improve the accuracy and reliability of fault diagnosis of high-voltage heaters, a fault monitoring method based on historical data of thermal power units is proposed. Firstly, the principal component analysis method is used to extract the 6-dimensional features that have a great influence on the outlet water temperature of the high-pressure heater, and the data is normalized. Then the XGBoost algorithm is used to establish a fault monitoring model, and the method is compared with the random forest and GBDT algorithms. Evaluate the quality of the model by mean absolute error (MAE), root mean square error (RMSE) and coefficient of determination (R2). Finally, the model is verified by the data of the high-pressure heater in the fault state. The results show that the PCA-XGBoost algorithm has the best effect in predicting the outlet water temperature of the high pressure heater.Through the comparison of the temperature prediction curve and the continuous fluctuation signal of the residual curve, the fault warning is realized. It shows that the fault monitoring model proposed in this paper can give early warning to the fault of the high-pressure heater, and can improve the reliability of the system operation.