This study addresses the limitations of existing construction site layout planning (CSLP) methods by proposing a dynamic multi-objective optimization model tailored to prefabricated construction projects. Unlike conventional static and phased approaches, the proposed model incorporates time-based facility placement, enabling more realistic and spatially efficient site layouts. It simultaneously optimizes key project objectives, including minimizing hoisting time, reducing hazard risk, and lowering transportation costs.
The model captures the temporal overlap of facilities and allows space reuse once a facility is no longer active. It incorporates a comprehensive set of spatial and operational constraints, such as crane working ranges, exclusion zones, overlap avoidance and site boundaries. To solve the model effectively, a hybrid metaheuristic algorithm, oMOAHA, is employed. This algorithm extends the original Multi-objective Artificial Hummingbird Algorithm (MOAHA) by integrating three opposition-based strategies. The model is validated through two real-world case studies: a private hospital and a feed manufacturing complex.
The dynamic approach consistently outperforms the static and phased methods by yielding Pareto-optimal solutions with superior quality. The oMOAHA algorithm also outperforms MOAHA and Non-dominated Sorting Genetic Algorithm II (NSGA-II) across multiple metrics, including hypervolume (HV), spacing (SP) and the number of Pareto-optimal solutions (NP). These results confirm the model's robustness for both bi-objective and tri-objective optimization in complex prefabricated environments.
This study introduces a novel dynamic CSLP model that advances the theoretical and practical understanding of site layout optimization. By combining time-based facility scheduling with a powerful hybrid optimization algorithm, the model fills a critical gap in the literature. Its validation through real-life case studies demonstrates practical applicability and highlights the potential for integration with digital technologies such as building information modeling (BIM) and Digital Twins to enhance adaptability and resilience in dynamic construction contexts.
