基于多机协同遗传算法自动化类结构设计

Design of Automation Class Structure Based on Multi-machine Collaborative Genetic Algorithm

摘要: 针对从需求到软件结构的自动化问题,提出基于多机协同遗传算法搜索最优软件结构的设计和基于耦合内聚量化指标新组合的适应度评价.通过实验表明,基于多机协同遗传算法比传统搜索算法收敛更快,新的适应度评价能更有效地指引搜索获取更优软件结构.

Abstract: To address the software automation problem from requirements to software structure, search of the optimal software structure based on multi-machine cooperative genetic algorithm (MCGA) is designed. Quality assessment index of cohesion and coupling as a new fitness function is proposed. Experiments show that the MCGA converges faster than traditional search algorithms and can get better software structure by the new fitness evaluation.