Optimal Individual Selection Algorithm Based on Layer Proximity and Branch Distance Functions

The article delves into the realm of automatic test case generation using heuristic algorithms, a critical area in ensuring software quality and reliability. It begins by discussing the current state of heuristic algorithms, including simulated annealing, genetic algorithms, and ant colony optimization, highlighting their strengths and limitations in solving combinatorial optimization problems. The core of the article introduces a novel algorithm that addresses the challenges in selecting optimal individuals during the iterative process. This algorithm leverages layer proximity and branch distance functions to enhance the comparability of individuals, thereby improving the navigation of populations towards optimal solutions. The article provides a detailed overview of the steps involved in automatic test case generation using the particle swarm algorithm, illustrating the importance of fitness function design and optimal individual selection. It also presents a comprehensive analysis of the problems associated with current individual selection methods, particularly in programs with deep nested structures. The proposed algorithm is thoroughly validated through four different experiments, demonstrating its superiority in terms of iteration count and execution time. The experimental results underscore the algorithm's effectiveness in generating test cases for complex programs, making it a valuable contribution to the field of automatic test case generation and heuristic optimization.