Simulink Model Static Analysis Results Based on Abstract Interpretation

Yuting Yang,Rui Wang,Youchen Wang,Xu Miao,Bing Liu,Shan Jiang

DOI: https://doi.org/10.1109/dsa56465.2022.00087

2022-01-01

Abstract:In the static analysis of Simulink models, there always exist false positives in the model rule checking results. Research shows that one of the important reasons is that the checking process ignores the model context information, which makes the program unable to make accurate judgments and warning potential risks. In response to this problem, researchers propose an optimization method based on model slicing and abstract interpretation to optimize the inspection results includes variable values or variable types. The optimization method first obtains the backward model slice that may affect the problem module as the focus of the analysis, converts the slice into a three-address code program, and constructs a control flow graph (CFG) after completing the basic block analysis of the program. At the same time, the abstract explanation theory is used to design and execute iterative algorithms for possible problems and constructed control flow graphs, infer variable values or variable types, and finally complete the optimization of the inspection results. In this paper, the effectiveness of our method is applied in two experiments and achieves decent performance.