Year: 2025
Type: conference-paper
Abstract: With the increasing complexity of safety-critical systems, the significance of testing these systems has grown correspondingly. Generating test cases based on the SysML models of systems offers an effective solution to address the challenges posed by system complexity, thereby improving the efficiency and thoroughness of testing. However, traditional symbolic execution techniques face limitations when dealing with black-box behaviors in SysML models. To overcome this, we propose an extension to SysML that allows modelers to explicitly identify black-box behaviors and indicate the parameters that may vary within these behaviors. Based on the extended SysML state machine model, we introduce an automatic test case generation algorithm grounded in symbolic execution. The generated test case suite achieves high Modified Condition/Decision Coverage (MC/DC). Our contributions include: 1) we propose an extension to SysML that supports the modeling of black-box behaviors, facilitating the automatic generation of test cases using symbolic execution methods; 2) we develop a symbolic execution-based test case generation algorithm that produces test case suites with high MC/DC; and 3) we validate the effectiveness of the proposed approach through experiments on the state machine of an automotive collision-avoidance system.
DOI: 10.1117/12.3062801
Authors: Chenghao Meng, Zhangtao Jia, Haoyu Li, Xiaoxiao Zhang, Jing Wang
Institution: Nankai University
Language: English
Subfield: Software
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4409785053