آيسان دين پژوه

Project management in construction—especially on a large scale—is a complex process that simultaneously requires meeting three key objectives: minimizing time, minimizing cost, an‎d maximizing quality. In practice, these three goals often conflict with one another, an‎d achieving an optimal balance among them is considered a major challenge for project managers. The presence of numerous activities, resource limitations, an‎d multiple contractors further complicates the establishment of such a balance. Many existing models examine these three criteria separately an‎d treat contractor selec‎tion as independent from scheduling, which leads to inconsistency in decision-making an‎d reduces the efficiency of outcomes. The research gap addressed in this study is the lack of an integrated, automated, multi-criteria decision-support framework, based on a metaheuristic algorithm, that can simultaneously selec‎t contractors while considering time constraints, cash-flow limitations, an‎d diverse, numerous precedence relationships, an‎d that can also provide various operational outputs. The research methodology is based on mathematical modeling an‎d the implementation of a modified genetic algorithm, which includes: (1) quantitative eva‎luation of contractor quality using the Analytic Hierarchy Process (AHP), (2) incorporation of financial an‎d temporal constraints, an‎d (3) generation of outputs such as Gantt charts, decision-support files, an‎d Pareto fronts. The algorithm performs contractor selec‎tion an‎d optimizes time, cost, an‎d quality simultaneously; ultimately, the project schedule is one of the key outputs of the proposed approach. Results from three case studies—an 18-activity benchmark problem, a three-objective road construction project, an‎d a real building construction project—demonstrated that the proposed model, compared with reference methods, achieves simultaneous improvements in time, cost, an‎d quality. For instance, in the classic 18-activity problem, the genetic algorithm proposed in this research successfully identified the complete an‎d optimal Pareto front in 110 days across all runs without requiring a time-limit constraint, whereas benchmark algorithms such as Particle Swarm Optimization (PSO) an‎d Ant Colony Optimization (ACO) failed to reach this time–cost performance in even 60% of their executions. Similarly, in the road construction case study, the superiority of the proposed algorithm over reference algorithms such as NSGA-II, MOPSO, OMODE, an‎d SMA was confirmed. In the real building project, the execution time decreased from 1425 to 1315 days (an 8% reduction) an‎d cost decreased from $691,188 to $690,928, while quality increased from 8.13 to 8.28. Overall, this study shows that the developed genetic algorithm can accurately achieve all objectives an‎d delivers better performance in complex projects—particularly where simultaneous multi-objective search is required. The core insight of this research is that integrating contractor selec‎tion an‎d scheduling within a unified multi-objective optimization framework can significantly improve construction project productivity an‎d provide project managers with diverse, reliable, an‎d practical decision alternatives.

بهينهسازي چندمعياره , انتخاب پيمانكار , الگوريتم ژنتيك , زمانبندي پروژه , كيفيت ساخت

Multi-criteria optimization , contractor selec‎tion , genetic algorithm , project scheduling , construction quality

Aysan Dinpazhooh

Mostafa Khanzadi