ScholarWorks@숙명여자대학교

초록

The rapid evolution of transportation cyber-physical systems (T-CPS) has led to unprecedented advancements in mobility and efficiency. However, ensuring the safety and reliability of these complex systems remains a critical challenge, particularly in verifying and refining their temporal behaviors. This paper presents a novel compositional approach for temporal behavior analysis and synthesis in safety-critical T-CPS, the safety-critical temporal analysis, and refinement for the T-CPS (STAR-TCPS) framework, which combines iterative compositional verification with L*-based learning techniques to analyze and refine timing behaviors efficiently. The method leverages the clock constraint specification language for high-level timing specifications and introduces a systematic refinement process to transform abstract requirements into implementable task models. The framework incorporates innovative constraint handling mechanisms tailored for T-CPS, addressing unique challenges such as vehicle-to-infrastructure communication delays and adaptive traffic management timing. Experimental evaluations, including a case study on an intelligent vehicle system, demonstrate STAR-TCPS's superiority over existing methods. Results show significant improvements in verification time (up to 63% reduction), memory usage (44% decrease), and task generation efficiency (18-25% fewer tasks) compared to state-of-the-art approaches. The STAR-TCPS framework enhances T-CPS's safety and reliability by enabling more efficient verification and refinement of critical timing properties, paving the way for safer and more robust transportation systems.

키워드