Optimizing co-simulation with the age of information in intelligent cyber physical systems

Abstract

Intelligent Cyber-Physical Systems (ICPS) integrate technologies from physics, information science, and artificial intelligence. Due to the specificity of application domains, these systems often demand higher standards for their quality. Utilizing co-simulation as one of the crucial techniques for system validation can effectively enhance system quality. However, due to the complexity of application scenarios, traditional co-simulation methods often lack precise descriptions of the temporal semantics in the interaction models, thereby overlooking the intricate temporal relationships among different components, resulting in the inability to flexibly, accurately analyze and evaluate system behavior based on different application scenarios. This paper introduces an ICPS co-simulation method that uses the Age of Information (AoI) mechanism to resolve identified problems. The co-simulation framework consists of physical simulation models together with information simulation models and interaction models which we designed first. We establish Age of Information-based temporal interaction types to precisely measure information freshness during decision-making processes. Three time synchronization protocols (time-stepping, global event-driven and variable-stepping) have been proposed to enhance simulator synchronization with an emphasis on operational efficiency and accuracy levels. Our approach validation includes running multiple simulation tests which examine different interaction types together with synchronization protocols. The modeling of information age of interaction results in enhanced decision-making accuracy according to experimental results. The proposed co-simulation method shows its practical feasibility based on RoboMaster EP experimental assessments which demonstrate similar behaviors between simulated and actual physical systems. The research enhances ICPS co-simulation techniques through information aging and synchronization solutions which lead to better system validation prior to deployment.