Research on multi-objective optimization of production line based on genetic algorithm
-
摘要: AGV(automatic guided vehicle)作为物料运输的载体,使生产线物料按节拍流动,是产线设计的关键部分。针对产线初步规划方案中AGV的数量、配送量及速度参数未确定影响产线最佳方案设计的问题,将产线产品生产周期、暂存区总容量、设备平均利用率及AGV平均利用率作为多个目标。采用全因子试验方法,探究AGV的数量、配送量及速度3个因素对多个目标的影响及其变化规律,确定关键因子与优化目标。建立多目标数学模型,采用遗传算法求解,调整并行工序数量,获得优化方案并仿真验证。研究结果表明:该方法能够有效求解问题,使作业的AGV数量减少至1辆,提高资源利用率,降低企业投资成本。Abstract: Automatic guided vehicle as the carrier of material transportation, make the production line material flow according to time, which is a key part of the production line design. Aiming at the problem that the AGV quantity, distribution quantity and speed parameters are not determined in the preliminary scheme of production line, which affects the optimal scheme of production line. The production cycle of production line, total capacity of temporary storage area of production line, average utilization rate of equipment and average utilization rate of AGV were taken as multi-objective. Full factor experiment method was used to study the influence of AGV quantity, distribution quantity and speed on multiple targets and its changing rules, identify key factors and optimization targets. A multi-objective mathematical model was established, and genetic algorithm was used to solve the problem, and adjust the number of parallel processes, the optimization scheme was obtained and verified by simulation. The results show that the method can solve the problem effectively, and reduce the number of working AGVs to 1, it improves resource utilization and reduces the investment cost of enterprises.
-
Key words:
- production line /
- experiment design /
- key factor /
- genetic algorithm /
- multi-objective optimization
-
表 1 工位信息
工位号 并行工序数量/个 加工产品数量/个 标准作业时间/s 工位1 1 1 73 工位2 1 1 35 工位3 4 4 51 工位4 1 1 72 工位5 1 1 27 工位6 1 1 38 工位7 1 5 240 工位8 4 4 180 工位9 8 8 900 工位10 1 1 40 表 2 全因子试验设计表
因子 下限 上限 增量 AGV数量x1 1 5 1 AGV配送量x2 5 10 1 AGV速度x3 0.5 1.5 0.1 表 3 最佳输出变量
因子 最佳输出变量 x1 x2 x3 Y1 Y2 Y3 Y4 1 9 1 503 67 *354 *0.513 *0.368 1 5 0.8 *531 21 308 *0.523 *0.454 2 5 1.3 *523 37 *332 0.532 *0.167 1 8 0.5 *515 21 *395 *0.511 0.572 注:带*数值为非最佳输出变量。 表 4 因子取值范围
因子 区间 x2 [5,10] x3 [0.5,1.5] 表 5 多目标最优解
因子 目标 x1 x2 x3 Y1 Y2 Y3 Y4 1 5 1 52642 322 0.528 0.397 表 6 二次优化与遗传算法优化
优化方法 工位资源利用率/(%) 工位3 工位8 遗传算法 10.57 37.31 二次优化 21.1 74.46 -
[1] 李慧, 孙元亮, 张超. 基于Plant Simulation的航空发动机叶片机加生产线仿真分析与优化[J]. 组合机床与自动化加工技术, 2019, 61(7): 116-118. [2] 李珍萍, 施莹, 吴凌云. 多约束混流线平衡与排序优化问题研究[J/OL]. 系统仿真学报. https://doi.org/10.16182/j.issn1004731x.joss.21-0679, 2022-03-20/2022-05-22. [3] 刘镇海, 周友行, 徐长锋. 基于Plant Simulation的路灯杆件生产线产能优化方案研究[J]. 现代制造工程, 2020, 43(1): 35-42. [4] Kassoul K, Cheikhrouhou N, Zufferey N. Buffer allocation design for unreliable production lines using genetic algorithmand finite perturbation analysis[J]. International Journal of Production Research, 2022, 60(10): 3001-3017. doi: 10.1080/00207543.2021.1909169 [5] Viharos A B, Németh I. Simulation and scheduling of AGV based robotic assembly systems[J]. IFAC PapersOnLine, 2018, 51(11): 1415-1420. doi: 10.1016/j.ifacol.2018.08.317 [6] 晁永生, 孙文磊, 袁亮. 平行装配线平衡性的多目标优化[J]. 计算机集成制造系统, 2016, 22(5): 1211-1219. doi: 10.13196/j.cims.2016.05.006 [7] 严金凤. 飞机总装脉动生产线数字化仿真与优化关键技术[J]. 航空制造技术, 2020, 63(20): 54-60. doi: 10.16080/j.issn1671-833x.2020.20.054 [8] 潘斌斌. 多目标路径规划问题的算法综述[J]. 重庆工商大学学报, 2012, 29(5): 78-84. [9] 李广振, 徐志刚, 任朝晖, 等. 规划阶段的自动化装配线仿真与优化[J]. 组合机床与自动化加工技术, 2021, 63(2): 157-160. doi: 10.13462/j.cnki.mmtamt.2021.02.038 [10] Malega P, Kadarova J, Kobulnicky J. Improvement of production efficiency of tapered roller bearing by using plant simulation[J]. International Journal of Simulation Modelling, 2016, 16(4): 682-693. doi: 10.2507/IJSIMM16(4)10.405 [11] 师路欢, 李耀辉, 吴义忠, 等. 序列Kriging仿真优化方法综述[J]. 机械设计与制造, 2021, 59(11): 279-286. doi: 10.3969/j.issn.1001-3997.2021.11.063 [12] 陈冠中, 陈庆新, 毛宁, 等. 智能制造系统AGV小车及缓冲区容量仿真优化[J]. 工业工程, 2016, 19(3): 77-84. doi: 10.3969/j.issn.1007-7375.2016.03.013 [13] 阳光灿, 熊禾根. 改进遗传算法求解柔性作业车间调度问题[J]. 计算机仿真, 2022, 39(2): 221-225. doi: 10.3969/j.issn.1006-9348.2022.02.042 [14] 张邦成, 夏奇, 鲁冠宏, 等. 针对车间平衡性能的工位布局优化研究[J]. 制造技术与机床, 2021, 71(10): 105-109. [15] 董海, 徐晓鹏, 谢谢. 多目标优化算法求解多柔性作业车间调度问题[J]. 计算机科学, 2020, 47(12): 239-244. doi: 10.11896/jsjkx.191100042 [16] 卢义桢, 李西兴, 朱传军, 等. 基于自适应遗传模拟退火算法的多目标车间布局优化[J]. 制造技术与机床, 2022(7): 173-179. doi: 10.19287/j.mtmt.1005-2402.2022.07.029