Combinatorial Testing (CT) is a widely used strategy in software testing that ensures coverage of parameter interactions using a minimal number of test cases, typically represented as Covering Arrays (CAs). A key challenge in generating effective CAs lies in the tendency of many search-based algorithms—especially meta-heuristic ones—to become trapped in local optima, leading to suboptimal solutions and high computational costs. Although various algorithms have been proposed to mitigate this issue, many still suffer from inefficiency and lack of consistency in producing high-quality test sets.
In this study, we introduce a novel meta-heuristic algorithm called Beautiful Mind (BM), inspired by the human cognitive process. The algorithm combines elements of both rational intelligence and emotional reasoning to guide the search process more adaptively and escape local optima. BM incorporates a dynamic adjustment mechanism to improve search diversity and convergence. Comprehensive experiments were conducted to evaluate its performance, showing that BM not only produces more optimal covering arrays but also achieves greater consistency and accuracy compared to existing meta-heuristic methods.