Search

Due to the introduction of large-scale renewable energy to the electric grid, the coal-fired units are running more under load cycling conditions and this has dramatically increased the difficulty of boiler in the control of NOx emissions. The real-time prediction of NOx emissions is of great significance to guide the efficient and clean operation of thermal power units under load cycling conditions. The CFD methods require huge computational cost to solve multiple coupled conservation equations, and hence, are infeasible to establish the complex nonlinear mapping relationship between boiler operating parameters and NOx emissions in a real-time manner. With the rapid development of big data analysis and artificial intelligence, data-driven models provide a new approach to realize the real-time prediction and control of boiler NOx emissions. Under load cycling conditions the boiler operating data has strong correlation in time sequence and complex spatial correlation characteristics. However, the traditional neural network methods cannot effectively capture these correlation characteristics, thus, greatly limiting their ability in the accurate prediction of boiler NOx emission. To resolve this problem, a boiler NOx prediction model based on the spatiotemporal attention graph convolutional network (AST-GCN) is proposed in this study. Not only can this model capture the spatial correlations among the boiler operating parameters, but also can it capture the dynamic relationship between the historical boiler operating data and NOx emissions. In addition, the attention mechanism embedded in the model can further improve the ability to adaptively extract the spatio-temporal correlation features from the boiler operating data, and increase the interpretability of the model. The prediction results based on the operating data of a 600 MW boiler demonstrate that the AST-GCN model exhibits a greatly improved prediction accuracy and generalization performance since it can effectively capture the spatial and time sequence correlations among the boiler operating parameters in comparison to the traditional neural network models.

Overheating and the resultant tube failures of boiler high temperature heating surfaces such as waterwall is one of the sore major problems affecting the safe operation of coal-fired power generating units. Boiler tube overheating generally occurs at some localized areas of boiler heating surfaces. To alleviate tube overheating and avoid the resultant tube failures, it’s necessary to monitor the tube temperature distribution in real-time to make preventive boiler operation adjustment. Due to the limitation of measurement method and the huge time cost of numerical method, currently there are still lack of methods that can realize the real-time prediction of waterwall tube temperature distribution during boiler operation. Therefore, this paper integrated the coupled heat transfer model with neural network and established a typical database of a 350 MW supercritical boiler using the coupled model, which contains 220 different typical boiler operating cases generated by adjusting 46 key operating parameters. Then, 4 400 expansion cases were derived at extremely time cost by the quick expansion method proposed in this paper. The deep neural network model was then constructed with 46 boiler operating parameters and waterwall coordinates as the model input and the tube temperature at the corresponding locations as the output. The DNN model was trained based on the comprehensive database containing 220 typical cases and 4 400 expansion cases. The MSE of the DNN model on validation set was only 0.005 3, the AUC5 was 0.988 and the calculation time was within 0.1 s. It illustrates that the data-driven model realized the real-time prediction of the waterwall tube temperature distribution over the entire boiler operating conditions by means of generalizing the numerical simulation results of finite cases. In addition, a quick database expansion method was proposed to address the problem that the heat transfer deterioration is difficult to be accurately predicted due to the insufficient data of the heat deterioration, and the prediction accuracy of the heat deterioration problem was significantly improved. Both the accuracy and the response speed of the model meet the demand of coal-fired plants for a real-time monitoring of tube temperature distribution.