考虑负载量均衡的自动拣货系统AGV任务分配优化

Optimization of AGV task assignment considering load balancing in RMFS

为提高AGV自动拣货系统作业效率、降低作业成本,在剖析造成系统拥堵的关键影响因素基础上,提出考虑负载量均衡的AGV任务分配双层规划模型,上层考虑总成本最小,下层通过构建多目标函数来最小化系统负载量标准差和AGV空闲率。针对传统GA求解任务分配问题效率低、易陷入局部最优等问题,提出了一种改进自适应遗传算法(SAGA),引入sigmoid函数用于适应度值的转换,并参与到自适应调整交叉变异算子的操作中,加入灾变策略防止算法出现早熟的情况。最后将该算法对不同规模算例进行仿真,结果证明,相比遗传算法、自适应遗传算法、自适应灾变遗传算法、蚁群算法,该改进算法在不同规模算例实验结果中均有明显优化,表明改进算法能更好地避免早熟、提升求解质量与收敛稳定性,有效地均衡了路网负载量,降低了作业总成本。

To enhance the efficiency of RMFS and reduce operating costs,this paper proposed a double-layer planning model for AGV task allocation considering load balancing,following an analysis of key factors causing congestion in RMFS.The upper layer sought to minimize the total cost,while the lower layer addressed the minimization of the system load standard deviation and AGV idle rate through a multi-objective function.It introduced a modified adaptive genetic algorithm(SAGA)to overcome the challenges of low efficiency and susceptibility to local optima in solving task allocation problems using traditional GA.It integrated the sigmoid function was to transform fitness values and participates in the adaptive adjustment of cross-mutation operators.Additionally,it incorporated catastrophic strategies to prevent premature convergence of the algorithm.Finally,it simulated the proposed algorithm on different scale cases,and the results demonstrate that,in comparison to the genetic algorithm,adaptive genetic algorithm,adaptive disaster genetic algorithm,and ant colony algorithm,the improved algorithm exhibits substantial optimization in the experimental results of various scale cases.This suggests that the enhanced algorithm effectively avoids premature convergence,enhances solution quality and convergence stability,actively balances road network load,and reduces total operating costs.