Generating Test Cases with 100-Percent Requirements Coverage Using Design of Experiments

Model-based testing has become a de facto methodology for flight-control law verification and validation today. The recent flight-control programs of commercial aircraft like the Boeing 747-8 and Airbus A350 XWB are verified and validated using model-based test methods. These projects are long-timeframe projects, and so modification and improvements in control laws are predictable. Test cases have to be generated for these modified segments, and the timeframes get compressed as the project approaches the certification phase. Hence, there is a need for test-case automation. Four methods of test-case automation are compared here for their efficacy in ensuring 100% requirements coverage and the time taken to generate such tests. A methodology of defining control theoretic requirements coverage is used here to generate the test cases. Random, simulated annealing, genetic algorithm, and design-of-experiments methods are used to generate the optimized set of test cases. The design of experiments is a method use...