| 摘 要: 针对NESMA(Netherl and Software Measurement Association)方法推荐权重不一定适合特定的项目群体及研究主要集中在软件开发工作量的局限性,使用蜉蝣算法对NESMA功能点组件的权重进行了调优。结合软件行业基准数据,提出了基于蜉蝣算法与 NESMA方法的一种全要素软件工程项目总投资估算方法。对该估算方法与遗传算法、粒子群优化算法、灰狼优化算法、模拟退火算法进行了实验对比,结果显示该方法将投资估算的精度优化0.22%~1.86%。此外,还对遗传算法的交叉进行了优化,提出多种交叉方法循环使用的方法,提高了交叉的效率。 |
| 关键词: 蜉蝣算法 NESMA方法 遗传算法 功能规模测量 软件开发成本估算 |
|
中图分类号:
文献标识码: A
|
| 基金项目: 河北省自然科学基金项目(2024209149);河北省省级科技计划资助项目(No6Z1019G) |
|
| Enhancing Precisionin Software Engineering Investment Estimation Using Mayfly Algorithm |
|
CAI Changshu1, KUANG Tianfu2
|
(1.Data Development & Promotion Center of Yunnan Province, Kunming 650051, China;
2.Chuxiong Normal University, Chuxiong Yi Autonomous Prefecture 675000, China)
490568995@qq.com; 3990184@qq.com
|
| Abstract: Addressing the limitations that NESMA method’s recommended weights may not suit specific project groups and existing research predominantly focuses on software development effort, this study employs the Mayfly Algorithm to optimize weights of Netherland Software Measurement Association(NESMA)functional components.Integrating software industry benchmark data, this study propose a comprehensive total investment estimation method for software engineering projects that accounts for all factors, combining the Mayfly Algorithm with the NESMA approach. Experimental comparisons with Genetic Algorithm(GA), Particle Swarm Optimization(PSO), Grey Wolf Optimizer(GWO), and Simulated Annealing Algorithm (SAA)demonstrate that the proposed method improves investment estimation accuracy by 0.22%-1.86% . Additionally, the crossover operation in Genetic Algorithms was enhanced by introducing a cyclical application of multiple crossover methods, thereby increasing crossover efficiency. |
| Keywords: mayfly algorithm NESMA method genetic algorithm functional size measurement softwaredevelopment cost estimation |
