Multi-objective optimization of railway construction under carbon trading mechanism based on improved NSGA-II algorithm

Abstract Railway engineering projects exhibit complex internal structures and extensive external interconnections, making it difficult to establish definitive control objectives for duration and cost during construction. With the increasing severity of environmental pollution and energy consumption, carbon emissions have gradually become a critical factor that cannot be ignored in the construction process. This study proposes a multi-objective optimization method for construction based on cost-duration-carbon emissions under the carbon trading mechanism, employing an improved NSGA-II algorithm to solve the multi-objective optimization model and deriving the optimal construction solution based on projection characteristic values. An empirical study of a railway engineering project demonstrates the following results: (1) Compared to the baseline solution, the optimized construction solution achieves reductions of 1.87 million yuan in cost, 3 days in duration, and 137 tons in carbon emissions, with all optimization results being statistically significant; (2) Quantitative analysis using spacing metrics and one-sample t-tests confirms that the improved NSGA-II algorithm exhibits high solution stability and solution set uniformity in solving construction multi-objective optimization problems, along with good scalability; (3) The optimal projection vector [Duration, Cost, Carbon Emission] = [0.0120, 0.9821, 0.9958] indicates that carbon emissions are the most critical factor influencing the selection of the optimal construction solution. This study provides theoretical foundations and technical support for multi-objective optimization in railway engineering construction, offering significant value for promoting green and sustainable development in the industry.