Multi-objective optimization of urea injection parameters of selective catalytic reduction system based SSABP-NSGA-II-TOPSIS

Selective catalyst reduction (SCR) system efficiency is significantly affected by issues like uneven ammonia distribution and incomplete urea decomposition, which are closely related to urea injection parameters. Initially, a 3D CFD model was developed to analyze how spray angle, offset angle, the distance between the nozzle and the front section of the catalyst, and injection velocity affect NOx conversion and NH3 uniformity. Then, a hybrid optimization framework combining SSABP neural network with NSGA-II and TOPSIS is used to optimize the impact of four parameters on the SCR system. The SSABP surrogate model showed good predictive accuracy compared with CFD simulations, with error metrics such as Mean Absolute Error (MAE) remaining within acceptable limits. The optimal injection parameters were identified as a spray angle of 35°, an offset angle of 2.5°, a nozzle-to-catalyst distance of 5.7 times the pipe diameter, and an injection velocity of 27 m/s. Under these conditions, the NOx conversion efficiency increased by 9.15 %, and the NH3 uniformity index improved from 0.71 to 0.845. The results demonstrate that the proposed SSABP–NSGA-II–TOPSIS framework is effective for optimizing urea injection strategies and enhancing SCR system performance under the investigated steady-state operating condition.