Integrated modeling of coking flue gas indices based on mechanism model and improved neural network

Focusing on the first domestic coking flue gas desulfurization and denitration integrated unit in China, the current condition of inlet flue gas indices cannot be determined timely owing to the large detection lag and complex upstream coking process, which is extremely unfavorable for the optimal control of desulfurization and denitration process. In order to solve this problem, an intelligent integrated modeling method of flue gas SO2 concentration, O2 content and NOx concentration is proposed. Firstly, the gas flow diagram in combustion process is built, the mechanism models of SO2, NOx concentration and O2 content are established according to the principle of material balance and reaction kinetics, respectively. Then the RBF neural network is adopted to compensate the prediction error, an improved training algorithm combining optimal stopping principle and dual momentum adaptive learning rate is proposed to improve the training speed and generalization ability of neural network. Based on the practical data of two 55-hole and 6-meter top charging coke ovens in the coking group, the effectiveness and superiority of proposed model and method are verified by simulation via comparison of various methods