Ant Colony Algorithm for Solving Multi-objective Optimization Problems in Software Engineering

With multi-objective becoming more and more important, ant colony algorithm has become an ideal tool to solve multi-objective optimization problems because of its unique positive feedback and distributed computing characteristics. This study explores the application of ant colony algorithm and its variants in multi-objective optimization problems of software engineering, especially the performance of ant colony system, and compares it with hybrid ant colony algorithm (HACO) and adaptive ant colony algorithm (AACO). The experimental results show that ACS is superior to AACO and HACO in convergence, execution time and solution set coverage. In particular, the maximum generation distance (GD) of ACS is only 0.2; The maximum execution time is 522ms, and the coverage rate is 80.3%-96.7%. These results confirm the outstanding performance of ACS algorithm in solving multi-objective optimization problems in software engineering.