Intelligent Construction Scheduling Based on MOEA/D-DE, SPEA2+SDE, and NSGA-III by Integrating Safety Assessment with Resource Efficiency

Abstract

To improve the efficiency and safety of intelligent construction scheduling, this work explores an optimization method for construction schedules based on multi-objective optimization (MOO) algorithms. This work focuses on the generation and optimization processes of scheduling plans and conducts safety assessments and resource efficiency analyses of the generated plans. The proposed optimized model is compared with classical MOO algorithms. These algorithms include Multi-Objective Evolutionary Algorithm based on Decomposition with Differential Evolution (MOEA/D-DE), Strength Pareto Evolutionary Algorithm 2 with Shift-based Density Estimation (SPEA2+SDE), and Non-dominated Sorting Genetic Algorithm III (NSGA-III). Based on the experimental results, the proposed optimized model outperforms three classic MOO algorithms across multiple key performance indicators. In terms of Hypervolume, the value achieved by the proposed model is 0.722, indicating that its solution set covers the objective space more effectively, demonstrating stronger diversity and global search capability. Furthermore, on the indicators of Generative Distance and Inverse Generative Distance, the proposed model attains lower values of 0.008 and 0.061, suggesting that the solution set is closer to the optimal front, with higher precision. In addition, the Spacing Metric value of 0.011 further shows that the solution set generated by the proposed model is more evenly distributed in the objective space. It avoids excessive clustering and enhances the uniformity and adaptability of the solutions. This uniformity is critical in practical construction scheduling optimization. This is because, under multiple conflicting objectives, a well-distributed solution set provides decision-makers with more options, enabling a better balance between safety and resource efficiency. Regarding safety assessment, Plan C has a high score of 4.63, indicating that under the optimization of the proposed model, the construction plan can achieve excellent performance in resource utilization and provide better safety guarantees. Similarly, Plan D, which demonstrates the highest resource efficiency, receives an overall score of 4.72, showcasing its outstanding advantages in resource usage and scheduling efficiency. These results validate the proposed model's applicability and flexibility under different constraints and objective functions.