利用改进粒子群优化的软件模块划分算法

Improved PSO-based Software Module Partition Algorithm

软件模块划分是软件工程领域一个重要并且非常复杂的问题,通过软件模块划分可以将规模庞大的复杂软件系统分解为易于理解和维护的子系统.针对现有软件模块划分算法收敛速度慢、划分效果不佳的问题,提出了一种基于复杂网络和群体智能算法的软件模块划分算法.首先对软件系统抽象形成复杂网络图,然后改进粒子群优化算法,使用最短路径初始化种群,使用概率选择的方式更新粒子位置,最后给出概率选择粒子群算法的全局收敛性证明.采用六个典型复杂软件项目实证结果表明,新算法稳定性更好,收敛速度更快,为软件模块划分问题提供了一个有效的工程化方法.

Software module partition is an important and complex problem in the software engineering,the large complex software system can be divided into some subsystems that are easy to understand and maintain through the software module partition. Aiming at solving the problem of slow convergence speed and the poor partition result,a software module partition algorithm based on complex network theory and swarm intelligence algorithm is proposed. Firstly,the algorithm converts the software system into complex network diagram,and then the particle swarm optimization algorithm is improved. The shortest path method is used to initialize the population and the probability selection method is used to update the particle positions. Finally,the global convergence proof of probabilistic selection particle swarm optimization is given. Experimental results of six typical complex software projects show that the algorithm is more stable and more convergent than traditional algorithm,which provides an effective engineering method for software module partitioning.