Hybrid simulation by combining system dynamics and agent-based modeling approaches in construction projects

Abstract

Hybrid modeling enables us to use strengths of various simulation approaches. System dynamics is a continuous simulation approach which uses feedback loops, stocks and flows to simulate the complicated behavior of complex systems over time. Agent based modeling is a simulation methodology which uses some specified rules to simulate the behavior of agents in their surrounding environment. Combining of system dynamics and agent-based modeling approaches enhance the capabilities and strengths of individual simulation paradigms. Also, it enables modelers to study more sophisticated problems. This paper presents a novel framework to integrate system dynamics and agent-based approaches to be implemented on construction projects. The proposed approach can provide a complete guideline for modelers by accounting for the most important issues which should be considered by modeler during integrating system dynamics and agent-based approaches. The framework proposes five steps to develop hybrid system dynamics and agent-based models. This step by step process helps to solve complex construction problems considering features of the problem. To evaluate the performance of the proposed approach it is implemented in a real project to investigate the unsafe behavior of different workgroups in a construction site taking account of the interactions with other working groups as well as the surrounding environment. Finally, the project duration is predicted taking account of unsafe behavior of different working groups.