摘要
针对自动仓储系统中多AGV的批量任务分配问题,以总任务等待时间、AGV负载均衡指数、总AGV能耗为目标,以AGV和任务的匹配关系为决策变量构建多目标优化数学模型,加入电量约束条件。为克服传统NSGA⁃Ⅱ算法存在的收敛速度慢、种群多样性维护差等不足,提出三种改进策略:改进交叉和变异算子,采用顺序交叉算子、逆序和单点相结合的混合变异算子;改进拥挤度计算公式,提出非线性平均绝对差的拥挤度计算方法;提出从局部和全局双角度调整的动态参数自适应调整策略。最后,设计多AGV多任务分配的仿真实验。实验结果表明:改进的NSGA⁃Ⅱ算法有效解决了多AGV多任务分配问题,同时,所提出的改进策略有效提高了算法的收敛速度、稳定性和鲁棒性。
To address the batch task allocation problem for multi⁃AGVs(automated guided vehicles)in an automated warehousing system,a multi⁃objective optimization mathematical model is formulated with the objectives of minimizing total task waiting time,optimizing AGV load balance index,and reducing overall AGV energy consumption.The matching relationship between AGVs and tasks is established as decision variables,incorporating electrical constraints.In order to overcome the shortcomings of the traditional NSGA⁃Ⅱ(non⁃dominated sorting genetic algorithmⅡ)algorithm,for example,slow convergence speed and poor maintenance of population diversity,three improvement strategies are proposed,including improving crossover and mutation operators and adopting a hybrid mutation operator combining sequential crossover operator,reverse order and single⁃point mutations,improving the crowding degree calculation formula by introducing a non⁃linear average absolute deviation method,and introducing a dynamic parameter adaptive adjustment strategy in both global and local perspectives.Simulation experiments for multi⁃AGV multi⁃task allocation are designed.Experimental results demonstrate that the improved NSGA⁃Ⅱalgorithm can effectively address the batch task allocation problem for multi⁃AGVs,and enhance the convergence speed,stability and robustness.
作者
王凡通
王凌
高雁凤
陈锡爱
王斌锐
WANG Fantong;WANG Ling;GAO Yanfeng;CHEN Xiai;WANG Binrui(College of Mechanical and Electrical Engineering,China Jiliang University,Hangzhou 310018,China)
出处
《现代电子技术》
北大核心
2024年第9期157-163,共7页
Modern Electronics Technique
基金
浙江省公益性技术应用研究(分析测试)计划项目(LGC21F030001)。
关键词
自动仓储系统
多AGV
任务分配
多目标优化
电量约束
动态参数
NSGA⁃Ⅱ算法
automated warehouse system
multi⁃AGV
task allocation
multi⁃objective optimization
electrical constraint
dynamic parameter
NSGA⁃Ⅱalgorithm
