A Test Suite Generation Approach Based on EFSMs Using a Multi-objective Genetic Algorithm
A Test Suite Generation Approach Based on EFSMs Using a Multi-objective Genetic Algorithm
复制标题
使用多目标遗传算法的基于 EFSM 的测试套件生成方法
DOI：
--
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
R. Lefticaru
中科院分区：
文献类型：
--
作者：
Ana Turlea;F. Ipate;R. Lefticaru
使用扩展的有限状态机生成测试数据可能是一个困难的过程，因为我们需要生成可行的路径，还需要找到遍历给定路径的输入数据。提出了一种扩展有限状态机的测试集生成算法。该算法使用一种改进的多目标遗传算法(删除冗余路径和缩短解)来产生一组覆盖所有过渡的可行过渡路径。该多目标优化问题的目标是基于数据流依赖关系对转移覆盖和路径可行性进行优化。有了这个算法得到的一组路径，我们可以很容易地找到每条路径的输入参数。
Using extended finite state machines for test data generation can be a difficult process because we need to generate paths that are feasible and we also need to find input data that traverse a given path. This paper presents a test suite generation algorithm for extended finite state machines. The algorithm produces a set of feasible transition paths that cover all transitions using a modified multi-objective genetic algorithm (deleting redundant paths and shortening the solutions). The multi-objective problem aims to optimize the transitions coverage and the path feasibility, based on dataflow dependencies. Having a set of paths resulted from this algorithm, we can easily find input parameters for each path.