摘要
围绕车辆调度问题多目标优化这一主题,依据企业现有的GPS定位系统,设计一种基于实数编码和精英保留策略,综合应用外部惩罚函数法的多目标遗传算法.并于其中创新性的引入基因校验算子,避免了重复车次的产生.程序使用Java语言编制,后台数据库使用oracle.实验结果表明,该算法能够充分满足系统硬约束条件,快速收敛于目标约束,完成车辆的优化调度,具有较强的实用价值.
This dissertation,base on GPS,design of a real-coded and elite retention strategies and the comprehensive application of external penalty function method for multi-objective genetic algorithm around the vehicle scheduling problem of multi-objective optimization.And also,in which the introduction of innovative gene validation operator to avoids repeat vehicle trips generated.Programs compiled using the JAVA language,database is ORACLE.Experimental result shows that the algorithm can fully accord with the system hardware constraints,optimal scheduling of vehicles and has a strong practical value.
出处
《小型微型计算机系统》
CSCD
北大核心
2010年第3期545-548,共4页
Journal of Chinese Computer Systems
关键词
遗传算法
车辆调度
GPS
惩罚函数
变异算子
genetic algorithm
vehicle scheduling
GPS
penalty function
mutation operator
