面向多AGV环境下的部分柔性作业车间调度方法研究

Research on Scheduling Methods for Partially Flexible Job Shops in Multi-AGV Environments

摘要:

针对现有调度方法在部分柔性作业车间调度问题下适应性差，并缺乏对多AGV物流调度考虑的问题，提出了一种基于贪婪方法的改进遗传算法。考虑了多AGV环境下的部分柔性作业车间调度问题，建立以加工机器资源有限约束、多AGV物流规则约束、最小化最大完工时间为目标的数学模型。提出一种改进遗传算法，通过三段式染色体编码策略来同时解决工序加工顺序、工序所选择的加工机器、工序所选择的物流运输设备三个子问题，并引入随机数优化变异和双保留策略产生并保留优秀个体编码，提高了算法的搜索精度和收敛速度。在解码中加入贪婪方法，从而进一步减少完工时间。以某五金加工企业的层篮板和上U管加工车间为例，采用改进算法进行车间调度优化，结果表明该方法可以在搜索空间中精准获取有效的调度方案，从而验证了方法的可行性和优越性。

Abstract:

Aiming at the poor adaptability of existing scheduling methods in partially flexible job shop scheduling problems and the lack of consideration for multi-AGV logistics scheduling, an improved genetic algorithm based on a greedy method is proposed. Considering the partially flexible job shop scheduling problem in a multi-AGV environment, a mathematical model is established with constraints of limited processing machine resources, multi-AGV logistics rules, and the objective of minimizing the maximum completion time. An improved genetic algorithm is proposed, which utilizes a three-stage chromosome encoding strategy to simultaneously address three subproblems: processing sequence of operations, machine selection for operations, and logistics transportation equipment selection for operations. The algorithm introduces random number-optimized mutation and a dual-retention strategy to generate and preserve high-quality individual codes, thereby enhancing search accuracy and convergence speed. A greedy method is incorporated into the decoding process to further reduce the completion time. Taking the layer guard plate and upper U-tube processing workshop of a metal processing enterprise as a case study, the improved algorithm is applied for workshop scheduling optimization. The results demonstrate that this method can accurately obtain effective scheduling solutions in the search space, verifying its feasibility and superiority.