首页 | 官方网站   微博 | 高级检索  
     

考虑学习/遗忘特性的软件项目调度多目标优化方法
引用本文:郭一楠,季俊华,吉建娇,巩敦卫. 考虑学习/遗忘特性的软件项目调度多目标优化方法[J]. 控制与决策, 2018, 33(2): 203-210
作者姓名:郭一楠  季俊华  吉建娇  巩敦卫
作者单位:中国矿业大学信息与控制工程学院,江苏徐州221008;中山大学机器智能与先进计算教育部重点实验室,广州510006,中国矿业大学信息与控制工程学院,江苏徐州221008,中国矿业大学信息与控制工程学院,江苏徐州221008,中国矿业大学信息与控制工程学院,江苏徐州221008
基金项目:国家自然科学基金项目(61573361);国家973计划项目(2014CB046300);中山大学机器智能与先进计算教育部重点实验室开放研究基金项目(MSC20170A);中国矿业大学创新团队项目(2015QN003).
摘    要:人员作为软件项目调度过程中的核心资源,其学习遗忘特性是无法忽视的.借鉴已有学习和遗忘模型,构建学习/遗忘效应与人员技能水平之间的动态关联模型,进而给出考虑人员学习/遗忘效应的软件项目调度多目标优化模型.针对该模型,采用新型调度方案编码方式和不可行解修复方法,给出基于改进NSGA-II的软件项目调度多目标优化方法.面向具有不同项目规模的算例仿真实验表明,考虑人员的学习能力有利于改善调度方案性能,而遗忘效应则会使调度方案的项目总工期和成本增加.因此,在软件项目调度问题中,考虑人员的学习和遗忘效应是十分必要的.

关 键 词:学习效应  遗忘效应  NSGA-II  多目标  软件项目调度

Multi-objective software project scheduling optimization method with the learning and forgetting effect
GUO Yi-nan,JI Jun-hu,JI Jian-jiao and GONG Dun-wei. Multi-objective software project scheduling optimization method with the learning and forgetting effect[J]. Control and Decision, 2018, 33(2): 203-210
Authors:GUO Yi-nan  JI Jun-hu  JI Jian-jiao  GONG Dun-wei
Affiliation:School of Information and Control Engineering,China University of Mining and Technology,Xuzhou 221008,China;Key Laboratory of Machine Intelligence and Advanced Computing of Ministry of Education,Sun Yat-sen University,Guangzhou510006,China,School of Information and Control Engineering,China University of Mining and Technology,Xuzhou 221008,China,School of Information and Control Engineering,China University of Mining and Technology,Xuzhou 221008,China and School of Information and Control Engineering,China University of Mining and Technology,Xuzhou 221008,China
Abstract:As a core resource in software project scheduling, employees'' learning and forgetting characteristic can not be ignored. Based on the existing learning and forgetting models, a model describing the dynamic relationship between the learning/forgetting effect and the employees'' skill levels is constructed. Then a multi-objective optimization model for software project scheduling considering above improved learning-forgetting model is given. The multi-objective software project scheduling optimization method based on the improved NSGA-II is introduced to solve this problem by using the new encoding mode and gene-based repair method for the infeasible offsprings. Experimental results under different project scales show that, considering the employees'' learning ability is beneficial to improve the performances of the optimal scheduling scheme, whereas the employees'' forgetting effect makes the duration and cost of the optimal scheduling scheme worse. Thus it is very necessary to give full consideration to the employees'' learning and forgetting effects in the software project scheduling problems.
Keywords:
点击此处可从《控制与决策》浏览原始摘要信息
点击此处可从《控制与决策》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号