基于群体智能算法的构件分类及选择方法研究

发布时间：2018-01-17 13:31

本文关键词：基于群体智能算法的构件分类及选择方法研究　出处：《湖南工业大学》2017年硕士论文　论文类型：学位论文

【摘要】：构件技术是支持软件复用的核心技术,如何从海量的构件库中选取出最优构件提高系统组装率,这是一个需要解决的关键问题。构件的选取一般包括两个步骤:构件的检索和构件的选择。构件的检索通常会选出多个构件类,效率是检索的核心问题;构件的选择是从构件检索的结果集中根据用户的需求选出构件,如何从检索出的构件类中选择可信性更高的构件是本文研究的另外一个问题。本文的主要研究工作如下:(1)针对构件检索效率的问题,首先对传统的构件分类技术做了分析,针对其局限性,在构件检索中引入构件挖掘分类的方法,利用蚁群分类算法对构件进行复用规则挖掘。其次针对蚁群算法易于早熟的问题对该算法做了改进,调整了蚂蚁的信息素浓度并引进了变异算子。最后通过实验验证改进的蚁群算法在构件查准率和查全率方面更有效。(2)针对构件选择可信性问题,本文从多个非功能属性对构件进行评估,采用粒子群算法选择满足用户需求的构件,根据该算法在选择构件的过程中出现“早熟”和易陷入局部较优的缺点,引入遗传算法的交叉和变异算子增加种群多样性,使得改进后的算法能够取得全局最优解。最后通过实验验证该算法有较好的收敛性,并且选择出的构件更满足用户的需求。
[Abstract]:Component technology is the core technology to support software reuse. How to select the optimal component from the massive component library to improve the system assembly rate. This is a key problem to be solved. The selection of components generally includes two steps: the retrieval of components and the selection of components. The retrieval of components usually selects multiple component classes, and efficiency is the core of retrieval. Component selection is to select components from the result set of component retrieval according to the needs of users. How to select a more credible component from the component class is another problem studied in this paper. The main research work of this paper is as follows: 1) aiming at the efficiency of component retrieval. Firstly, the traditional component classification technology is analyzed, aiming at its limitation, component mining classification is introduced in component retrieval. Ant colony classification algorithm is used to mine the reuse rules of components. Secondly, the algorithm is improved to solve the problem that ant colony algorithm is easy to prematurity. The pheromone concentration of ants is adjusted and mutation operator is introduced. Finally, the improved ant colony algorithm is proved to be more effective in component recall and recall rate. In this paper, the component is evaluated from several non-functional attributes, and the particle swarm optimization algorithm is used to select the components that meet the user's needs. According to this algorithm, the components are precocious and prone to fall into local optimization. The crossover and mutation operators of genetic algorithm are introduced to increase population diversity, so that the improved algorithm can obtain the global optimal solution. Finally, the experimental results show that the algorithm has better convergence. And the selected components more meet the needs of users.
【学位授予单位】：湖南工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TP311.5

【参考文献】