Multi-objective optimization of the combustion characteristics of moisture and soil containing wheat stubble by RSM and NSGA-II

• Quantified the inhibitory effects of moisture content and rhizosphere soil content on wheat stubble combustion. • Identified optimal excess air ratio and critical thresholds to maintain combustion efficiency. • RSM and NSGA-II are combined for the multi-objective optimization of the combustion-emission performance. To address the issue of soil-borne pathogens accumulating in crop stubble, this study conducted combustion experiments on high-moisture, soil-attached wheat stubble to provide a technical basis for in-field harmless disposal of residues and the prevention of soil-borne diseases. Thermogravimetric analysis (TGA) and a self-designed dual-grate reverse combustion system were employed to investigate the effects of excess air ratio ( λ ), moisture content (MC), and rhizospheric soil content (RSC) on combustion characteristics and pollutant emissions. A response surface model was developed using the Box-Behnken design, and the statistical importance of each is assessed by means of the analysis of variance (ANOVA), and the models are expressed in quadratic polynomial forms. A comprehensive comparison was conducted among the improved constrained multi-objective particle swarm optimization (CMOPSO), non-dominated sorting genetic algorithm II (NSGA-II), genetic algorithm (GA), and particle swarm optimization (PSO). The NSGA-II algorithm demonstrated superior performance in terms of Pareto front diversity, combustion efficiency, and control of NO x and particulate matter (PM) emissions. Based on this, multi-objective optimization was performed using NSGA-II, yielding the optimal operating conditions ( λ = 1.64, MC = 8.54%, RSC = 3.09%), which achieved a combustion efficiency of 85.43% while maintaining NO x and PM emissions at 99.25 mg/m 3 and 184.31 mg/m 3 , respectively, providing valuable guidance for sustainable energy utilization and environmental management.