A Combination Test Suite Generation Method Based on Adaptive Simulated Annealing Genetic Algorithm for Software Product Line Testing

Abstract

A Based-Similarity Combination SPL Test Suite Generation Method can be used to compute coverage, replace poor test cases and generate high coverage test suite. Improve the way to compute mutation and make test suite generation adapt to the current coverage from great mutation, the optimization procedure can be more efficient. Meanwhile, adaptive simulated annealing algorithm combined with simplified GA algorithm, which can ensure local optimization accuracy and take the whole situation into global searching space accordingly, is a good choice for combinatorial optimization problems. Experiments with 6 feature model from SPLOT shows that the new test suite generation method can achieve smaller-size test suite with higher coverage.