计算机科学
帧(网络)
软件
代码覆盖率
基于模型的测试
任务(项目管理)
认证
验证和确认
回归检验
测试线束
编码(集合论)
可靠性工程
测试用例
软件工程
软件建设
软件系统
程序设计语言
系统工程
工程类
集合(抽象数据类型)
机器学习
运营管理
电信
法学
回归分析
政治学
作者
Chandrashekhar Singh,Jagadish Shivamurthy,Asha Rani Garg
标识
DOI:10.1109/iccece51049.2023.10085332
摘要
Verification and Validation (V&V) of Safety Critical Software such as that of Flight Control Computer is an effort intensive and critical task. Accordingly standards like DO-178C lay much importance on the V&V aspects of the airborne software. Generation of typical and relevant test vectors over short sequences of time, calculating the expected outputs on a frame-by-frame basis and thus confirming the behavior of the software are daunting tasks. This paper brings out a Model Based Test (MBT) Framework to accelerate the functional testing. The Framework leverages the fact that, key and functional aspects of flight control such as Control Laws including advanced features such as Auto pilot are modeled and validated through simulation studies before generating software requirements. The Framework provides GUI based utilities to generate test vectors involving several external inputs over time durations such as 10 seconds covering multiple requirements. Such vectors are converted in to frame-by-frame test matrix which is fed to both the model and the program under test. The outputs of the model and program are automatically compared and differences beyond threshold are flagged for examination. The Framework additionally generates structural coverage reports both at the model and code levels which are the means to figure out that not only 100% MC/DC for the code is achieved but also there are no missing requirements. These coverage reports provide confidence and certification credits.
科研通智能强力驱动
Strongly Powered by AbleSci AI