A framework for conceptualising hybrid system dynamics and agent-based simulation models

Abstract

The growing complexity of systems and problems that stakeholders from the private and public sectors have sought advice on has led systems modellers to increasingly use multimethodology and to combine multiple OR/MS methods. This includes hybrid simulation that combines two or more of the following methods: system dynamics (SD), discrete-event simulation, and agent-based models (ABM). Although a significant number of studies describe the application of hybrid simulation across different domains, research on the theoretical and practical aspects of combining simulation modelling methods, particularly the combining of SD and ABM, is still limited. Existing frameworks for combining simulation methods are high-level and lack methodological clarity and practical guidance on modelling decisions and elements specific to hybrid simulation that modellers need to consider. This paper proposes a practical framework for developing a conceptual hybrid simulation model that is built on reviews and reflections of theoretical and application literature on combining methods. The framework is then used to inform and guide the process of conceptual model building for a case study in controlling the spread of COVID-19 in care homes. In addition, reflection on the use of the framework for the case study led to refining the framework itself. This case study is also used to demonstrate how the framework informs the structural design of a hybrid simulation model and relevant modelling decisions during the conceptualisation phase.