An Integrated Architecture for Validation of Simulation Models and Validation Methods Classification

Abstract

Validation of simulation models is a crucial aspect of the software engineering simulation modelling process. It aims to reduce costs, save time, mitigate risks, and enhance understanding of the validation process’s functions. This paper reviewed existing literature on simulation model validation, focusing on definitions, architectures, and methods. However, the study identified that it remains unclear which attributes should be prioritized when validating simulation models.  Furthermore, the literature lacks a comprehensive architecture and classification to effectively describe the validation process. To address these gaps, this paper proposes an integrated architecture for the Simulation Model Validation Process (SMVP). Furthermore, it introduces a novel classification that incorporates three essentials elements: simulation validation phases, simulation models, and validation methods. The findings indicate that the primary focus of verification method is to verify the accuracy of model behavior. In contrast, structural validation emphasizes historical data validation, historical methods, and internal validation. Moreover, combining the structural and behavioral methods ensure a more robust and effective validation process. In conclusion, the proposed architecture and classification are expected to significantly contribute to future research by ensuring that simulation models are thoroughly validated during the validation processes.