Software testing is crucial in verifying that system functionality meets specified requirements, especially in safety-critical railway control applications. This research employs Model-Based Testing to automate the generation of checking sequences for the Automatic Power Change-over system, leveraging a Python-based tool to create and refine test sequences from formalized UPPAAL models. Our approach integrates a clonal selection algorithm to inject diversity into the searching process, along with a reconstruction technique grounded in breadth-first search that preserves beneficial coverage paths, preventing coverage regression and ensuring more systematic exploration of complex state spaces. Experimental results show that transition coverage can be improved from an initial range of 50–60% to full transition coverage through iterative enhancements, requiring fewer refinements compared to purely random methods. This combination proves particularly effective for the Automatic Power Change-over system, which attained complete coverage with fewer iterations. These findings underscore the utility of structured mutation strategies and parameter optimization for achieving comprehensive coverage. Future work will investigate the integration of genetic algorithms to balance diversity and efficiency, as well as adopting more extensive coverage criteria suitable for train control systems. Converting these optimized checking sequences into executable test cases will further enhance safety and reliability in real-world operations.
Liao et al. (Tue,) studied this question.