An artificial intelligence optimization of NOx conversion efficiency under dual catalytic mechanism reaction based on multi-objective gray wolf algorithm

In the era of industry 4.0, artificial intelligence (AI) offers new perspectives for researching the complex sustainable chemical reactions in selective catalytic reduction (SCR). This aims to further improve the utilization and efficiency of SCR. In this study, a fuzzy gray relational analysis coupled with random forest (RF) and back propagation artificial neural network (BP-ANN) model was developed. This model was trained based on the Langmuir-Hinshelwood and Eley-Rideal coupled mechanism for SCR reaction mechanism, and had good fitting effect on the heat transfer rate, catalytic efficiency and ammonia (NH 3 ) slip rate of the catalytic reaction under loading conditions. And this was used as a guiding method to direct the multi-objective gray wolf optimization algorithm to optimize the basic parameters. The optimization results showed that the NH 3 slip rate of the SCR was slightly improved and the denitrification efficiency was increased up to 28 % under different loads, which had guiding significance for the lightweighting and thermal control of industrial equipment. • AI model is employed to investigate complex NO x conversiton for multi-objective optimization. • Langmuir-Hinshelwood and Eley-Rideal are coupled for SCR reaction mechanism. • The multi-objective gray wolf optimization algorithm is developed and optimized.