TACS车载设备复杂运营场景分阶段组合测试方法

Staged combinatorial testing method for complex operational scenarios of TACS system on-board equipment

基于车车通信的TACS(Train Autonomous Control System)列车运行系统通过深度融合车载控制系统和信号系统来实现线路资源管理、列车定位及追踪运行等各项关键功能,对可靠性和安全性的要求严格。针对TACS系统在复杂场景下的车载设备测试存在输入空间组合爆炸、测试用例冗余甚至无效等问题,提出一种车载设备分阶段组合测试方法。首先,以TACS车载设备自动折返场景为研究对象,依据相关规范文件分析其功能需求,提出测试阶段的划分原则与方法,辨识各个阶段下的测试参数、取值及约束,建立各阶段测试模型。其次,提出一种基于测试需求分阶段优化覆盖的组合测试用例生成算法,通过计算各阶段下的测试需求,采用基于贪婪覆盖策略的算法,结合Z3约束求解器,动态生成组合测试用例,实现约束异构影响下的测试需求全覆盖。最后,得到满足约束条件和覆盖强度的车载设备测试用例集。研究结果表明:该方法能够生成满足约束及覆盖率的测试用例集,有效减少了同一场景下模型间67.75%的冗余覆盖;与主流测试工具进行对比,该方法在覆盖强度为2时减少约50%的冗余测试用例,在覆盖强度为3时减少约36%的冗余测试用例。研究成果通过提高测试用例集的精简性,有效地降低了测试成本,为TACS车载设备测试用例的生成提供了参考。

The Train Autonomous Control System(TACS)is a train operating system that integrates the on-board control system and signaling system.It is based on vehicle-to-vehicle communication,fulfilling several important functions,including line resource management,train location,and tracking operation.The reliability and safety must be strictly satisfied.In response to the issues of input space combinations explosion,redundancy and even invalid tests when testing on-board equipment of TACS in complex scenarios,a staged combinatorial testing method was proposed.First,by taking as case study the automatic turn-back scenario,the stage division principles were proposed based on the analysis of functional requirements towards system specification.The test models were created for each stage by identifying the test parameters,values,and constraints.Second,a test generation algorithm was proposed,which covered the test requirements by stages in an optimal way.The algorithm generates tests by first computing the test requirements under each stage and then adopting a greedy generation strategy that covers the most test requirements in a stage.The Z3 constraint solver was used for valid check during test generation.Finally,the tests were generated by the proposed method that satisfies the constraints and coverage.The results show that compared to the popular test tools,the method can generate test case sets effectively,reducing 67.75%redundant coverage among models in the same scenario.50%fewer tests are generated at a strength of 2,and 36%fewer tests are generated at a strength of 3.Thus,the proposed method provides guidance for testing TACS on-board equipment with fewer tests and higher efficiency.