Performance of global optimization models for dynamic site layout planning of construction projects

Dynamic construction site layout planning is a complex optimization problem that is characterized by nonlinear objectives and constraints, which impose great challenges in obtaining global and feasible solutions. This paper presents and compares between two global optimization models of dynamic site layout planning that were developed to overcome the limitation of previous models in the literature. The first model utilizes Genetic Algorithms (GA) while the second model utilizes Approximate Dynamic Programming (ADP). The performance of these two optimization models is analyzed in terms of the effectiveness of reaching optimum solutions and the efficiency of reducing the computational time. This analysis is performed using a designed set of problems of dynamic site layout planning with changing size and complexity. It was found that ADP outperformed GAs in terms of effectiveness and efficiency. However, GAs still prove to be a viable optimization tool because of its simplicity and multi-objective optimization capabilities.