IEEE Access (Jan 2024)

Formal Verification Method of CTCS-2 Level Train Control Engineering Data Based on the Reduced Ordered Binary Decision Diagram

  • Hao Zhang,
  • Qing Xu,
  • Ke Ye

DOI
https://doi.org/10.1109/ACCESS.2024.3435468
Journal volume & issue
Vol. 12
pp. 106793 – 106807

Abstract

Read online

The precise generation of train control curves for the on-board Automatic Train Protection (ATP) of the Chinese Train Control System Level 2 (CTCS-2) relies significantly on accurate train control engineering data, which serves as a critical element in ensuring the safe operation of trains. However, the traditional verification methods of train control engineering data depend on manual validation, which lacks timeliness and completeness and makes it easy to overlook potential errors. On the other hand, traditional verification rules are derived from railway technical specifications and standards, expressed in textual language that is often ambiguous, which leads to insufficient completeness in data verification. To address these challenges, this paper proposes a formal verification method for train control engineering data based on a Reduced Ordered Binary Decision Diagram (ROBDD). First, the attribute constraints of the train control data and the implicit constraint relations between different data objects are mined using existing railway technical specifications and expert knowledge of railway signals. These constraints are then converted into Boolean function models. Second, we formulate the ROBDD generation algorithm and the evaluation decision algorithm of the Boolean function using the equivalent canonical data structure ROBDD of the Boolean function. Finally, based on the train control engineering data from actual passenger-dedicated lines, the automatic verification method is developed by constructing four types of “mutation” data, including mutate, swap, add, and remove. The test cases indicate that our proposed formal verification method is feasible and capable of achieving high efficiency and completeness in the verification of CTCS-2 level train control engineering data.

Keywords