基于模型的嵌入式软件自动化测试研究与实现

Research and Implementation of Model-Based Automated Testing for Embedded Software

摘要:

针对装备嵌入式软件规模扩增、接口交联复杂、版本迭代频繁，传统文档驱动测试存在标准化程度低、测试效率不足、异常场景覆盖不全的工程问题，提出一种改进型模型驱动嵌入式软件自动化测试方法。依托基于模型的系统工程（MBSE, Model-Based Systems Engineering）理论，采用逻辑-数据解耦的优化统一建模语言（UML, Unified Modeling Language）序列图建模方式，构建常规与异常全覆盖的测试模型，通过测试模型自动分解、测试数据与测试脚本自动生成、自动化测试执行及测试文档自动输出的全流程自动化方案，打通需求模型、测试模型、测试用例与测试脚本之间的技术壁垒，实现测试流程一体化集成。工程应用验证表明，该方案可有效提升嵌入式软件测试的数字化与标准化水平，显著优化测试效率与测试质量，适配复杂装备嵌入式软件频繁迭代的测试应用需求。

Abstract:

Equipment embedded software features expanding scale, intricate interface couplings and frequent version iterations. Traditional document-driven testing therefore suffers from poor standardization, low test efficiency and incomplete coverage of abnormal scenarios. This paper proposes an improved model-driven automated testing technology for embedded software. Based on Model-Based Systems Engineering (MBSE), an optimized Unified Modeling Language (UML) sequence diagram modeling approach with logic-data decoupling is adopted to build test models covering both conventional and abnormal scenarios. A full-process automated testing framework is established, supporting automatic decomposition of test models, automatic generation of test data and scripts, automated test execution and auto-generation of test documents. This framework eliminates technical silos between requirement models, test models and test cases as well as test scripts, and achieves seamless integration of the entire testing workflow. Engineering application results demonstrate that the proposed method effectively elevates the digitalization and standardization of embedded software testing, improves test efficiency and test quality, and can well adapt to iterative testing requirements of embedded software for complex equipment.