Towards a self-adaptive agent-based simulation model

Agent-based simulation (ABS) modelling has been a widely applied approach for simulating domain-specific phenomena. Currently, parameters and environments are simulated by a domain-specific model that is strictly used proprietarily by the ABS model developer. This causes inflexibility towards extension of the developed ABS model, which will further result in difficulties for validation and verification of the robustness and reliability of the ABS model. To address this issue, this paper proposes a self-adaptive ABS model that is capable of modelling cross-domain phenomena by selecting the required parameters based on the environment. The capability to self-adapt will allow the model to be easily extended and replicated. The self-adapt capability is enabled by a governing algorithm within the model and is conceptually illustrated through a case study of crime report process ABS modelling.