Abstract: Faced with the dual constraints of global resources and environmental sustainability, along with strategic demands for a national sustainable energy transformation, thermal power units are urgently required to undergo a lowcarbon transition to achieve carbon peak and carbon neutrality goals. Propose a two-stage coal blending strategy optimization method integrating carbon emissions and cost control for the high proportion of carbon emissions and cost control requirements of thermal power units. The aim is to strike an optimal balance between reducing environmental impacts and controlling costs, thereby effectively addressing the policy and market challenges facing thermal power units in the new era. Firstly, an improved adaptive chaos Non-dominated Sorting Genetic Algorithm II（NSGA-II） is used for the biobjective optimization of the coal blending strategy for thermal power units, focusing on the trade-off between carbon emissions and coal blending costs. A series of environmentally friendly and economical coal blending schemes that meet the actual operational constraints are determined through the obtained Pareto optimal solution set. Further analysis of the Pareto optimal set is conducted using the VIKOR method, comparing multiple evaluation criteria to rank the coal blending strategy options. This ensures that the decision-making process is systematic and diverse, providing power plant managers with clear and actionable coal blending strategy choices and offering innovative and practical solutions for low-carbon operation of thermal power units.This method not only enhances the efficiency of coal quality utilization and the eco-friendliness of thermal power units but also paves new pathways for the coal power industry to conform to national ’dual carbon’ targets and transition towards a future of low-carbon power generation, possessing significant theoretical and practical value.