Modeling and optimization of the NOx emission characteristics of a tangentially fired boiler with artificial neural networks

The present work introduces an approach to predict the nitrogen oxides (NOx) emission characteristics of a large capacity pulverized coal fired boiler with artificial neural networks (ANN). The NOx emission and carbon burnout characteristics were investigated through parametric field experiments. The effects of over-fire-air (OFA) flow rates, coal properties, boiler load, air distribution scheme and nozzle tilt were studied. On the basis of the experimental results, an ANN was used to model the NOx emission characteristics and the carbon burnout characteristics. Compared with the other modeling techniques, such as computational fluid dynamics (CFD) approach, the ANN approach is more convenient and direct, and can achieve good prediction effects under various operating conditions. A modified genetic algorithm (GA) using the micro-GA technique was employed to perform a search to determine the optimum solution of the ANN model, determining the optimal setpoints for the current operating conditions, which can suggest operators’ correct actions to decrease NOx emission.