Sequencing of Mixed Model Assembly Lines Based on Improved Shuffled Frog Leaping Algorithm

Shuffled frog leaping algorithm( SFLA) was used to solve multi-objective sequencing problem of mixed model assembly line( MMAL). Local convergence can be avoided and optimal solution can be obtained to a certain extent. However,the multi-objective sequencing problem of MMAL is an non-deterministic polynomial hard( NP-hard) problem and the shortcomings are slow convergence rate and low precision. To solve the shortcomings for optimization objectives of minimizing total utility time and keeping average consumption rate of parts, a chaos differential evolution SFLA( CDESFLA) is proposed in this study. Because SFLA is easy to fall into local optimum,the evolution operator of differential evolution algorithms is introduced in SFLA as a local search strategy,and differential mutation operator is introduced in chaotic sequence to prevent premature convergence. The examples show that the proposed CDESFLA is better for convergence accuracy than SFLA,genetic algorithm( GA) and particle swarm optimization( PSO)

Heuristics for Mixed Model Assembly Line Balancing Problem with Sequencing

The growing global competition compels organizations to use many productivity improvement techniques. In this direction, assembly line balancing helps an organization to design its assembly line such that its balancing efficiency is maximized. If the organization assembles more than one model in the same line, then the objective is to maximize the average balancing efficiency of the models of the mixed model assembly line balancing problem. Maximization of average balancing efficiency of the models along with minimization of makespan of sequencing models forms a multi-objective function. This is a realistic objective function which combines the balancing efficiency and makespan. This assembly line balancing problem with multi-objective comes under combinatorial category. Hence, development of meta-heuristic is inevitable. In this paper, an attempt has been made to develop three genetic algorithms for the mixed model assembly line balancing problem such that the average balancing efficiency of the model is maximized and the makespan of sequencing the models is minimized. Finally, these three algorithms and another algorithm in literature modified to solve the mixed-model assembly line balancing problem are compared in terms of the stated multi-objective function using a randomly generated set of problems through a complete factorial experiment.

Genetic Algorithm for Concurrent Balancing of Mixed-Model Assembly Lines with Original Task Times of Models

The growing global competition compels manufacturing organizations to engage themselves in all productivity improvement activities. In this direction, the consideration of mixed-model assembly line balancing problem and implementing in industries plays a major role in improving organizational productivity. In this paper, the mixed model assembly line balancing problem with deterministic task times is considered. The authors made an attempt to develop a genetic algorithm for realistic design of the mixed-model assembly line balancing problem. The design is made using the originnal task times of the models, which is a realistic approach. Then, it is compared with the generally perceived design of the mixed-model assembly line balancing problem.

Solving Level Scheduling in Mixed Model Assembly Line by Simulated Annealing Method

This paper presents an application of the simulated annealing algorithm to solve level schedules in mixed model assembly line. Solving production sequences with both number of setups and material usage rates to the minimum rate will optimize the level schedule. Miltenburg algorithm (1989) is first used to get seed sequence to optimize further. For this the utility time of the line and setup time requirement on each station is considered. This seed sequence is optimized by simulated annealing. This investigation helps to understand the importance of utility in the assembly line. Up to 15 product sequences are taken and constructed by using randomizing method and find the objective function value for this. For a sequence optimization, a meta-heuristic seems much more promising to guide the search into feasible regions of the solution space. Simulated annealing is a stochastic local search meta-heuristic, which bases the acceptance of a modified neighboring solution on a probabilistic scheme inspired by thermal processes for obtaining low-energy states in heat baths. Experimental results show that the simulated annealing approach is favorable and competitive compared to Miltenburg’s constructive algorithm for the problems set considered. It is proposed to found 16,985 solutions, the time taken for computation is 23.47 to 130.35, and the simulated annealing improves 49.33% than Miltenberg.

Mixed-Model U-Shaped Assembly Line Balancing Problems with Coincidence Memetic Algorithm

Mixed-model U-shaped assembly line balancing problems (MMUALBP) is known to be NP-hard resulting in it being nearly impossible to obtain an optimal solution for practical problems with deterministic algorithms. This paper pre-sents a new evolutionary method called combinatorial optimisation with coincidence algorithm (COIN) being applied to Type I problems of MMUALBP in a just-in-time production system. Three objectives are simultaneously considered;minimum number workstations, minimum work relatedness, and minimum workload smoothness. The variances of COIN are also proposed, i.e. CNSGA II, and COIN-MA. COIN and its variances are tested against a well-known algo-rithm namely non-dominated sorting genetic algorithm II (NSGA II) and MNSGA II (a memetic version of NSGA II). Experimental results showed that COIN outperformed NSGA II. In addition, although COIN-MA uses a marginal CPU time than CNSGA II, its other performances are dominated.

Minimizing Products Rates Variation in Just-in-Time Mixed-Model Manufacturing

Mixed-Model assembly lines are often used in manufacturing based on just-in-time techniques. The effective utilization of these lines requires a schedule for assembling the different models be determined. The objective is to minimize the total deviation of actual production rates from the desired production rates. Mathematical method with the optimization algorithm is proposed here to solve this problem. To prove the efficiency of the proposed algorithm, a number of test problems are solved. The results show that the proposed algorithm is an efficient and effective algorithm which gives better results with the large problem sizes. This paper presents a practical procedure to minimize total product variation rates, and easy to use by practitioner.

机器人人工拣选环境下混流装配线齐套物料配送优化

为了提高汽车制造企业配送的智能化水平、解决混流装配线齐套物料人工拣选效率低、成本高的问题,引入基于齐套物料配送策略的机器人人工拣选模式,通过优化自动拣选机器人、工人的配置数量及配送周期,使包含机器人使用成本、劳动力成本和在制品库存成本在内的总成本最小。为求解该配送优化问题,提出了改进量子蚁群算法,利用量子比特的叠加性增加种群多样性,避免算法陷入局部最优,同时设计了改进量子旋转门更新机制和基于差分进化操作的非最优个体优化策略,提高了算法收敛速度和寻优质量。最后,通过算例分析验证了模型的正确性与算法的有效性,并分析了拣选批量对总成本的影响。

Strategic Robust Mixed Model Assembly Line Balancing Based on Scenario Planning

Assembly line balancing involves assigning a series of task elements to uniform sequential stations with certain restrictions. Decision makers often discover that a task assignment which is optimal with respect to a deterministic or stochastic/fuzzy model yields quite poor performance in reality. In real environments, assembly line balancing robustness is a more appropriate decision selection guide. A robust model based on the α worst case scenario is developed to compensate for the drawbacks of traditional robust criteria. A robust genetic algorithm is used to solve the problem. Comprehensive computational experiments to study the effect of the solution procedure show that the model generates more flexible robust solutions. Careful tuning the value of α allows the decision maker to balance robustness and conservativeness of as- sembly line task element assignments.

Part Supply Method for Mixed-Model Assembly Lines with Decentralized Supermarkets

In-house part supply affects the efficiency of mixed-model assembly lines considerably. Hence, we propose a reliable Just-In-Time part supply strategy with the use of decentralized supermarkets. For a given production sequence and line layout, the proposed strategy schedules tow train routing and delivery problems jointly to minimize the number of employed town trains and the traveling time, while ensuring that stations never run out of parts. To solve this problem, a mathematical formulation is proposed for each sub-problem aiming at minimizing supply cost. Then, a dynamic programming algorithm for routing and a greedy algorithm for delivery are developed, both of which are of polynomial runtime. Finally, a computational study is implemented to validate the effectiveness of the strategy, and to investigate the effects of the delivery capacity of tow trains and storage capacity of stations on supply cost.

AN OPTIMAL KANBAN SYSTEM IN A MULTI-STAGE,MIXED-MODEL ASSEMBLY LINE

This paper studies the parameter design and the performance optimization of a Kanban system without stockouts in a multi-stage, mixed-model assembly line. The model of a Kanban system based on production processes is established by examining the relationship among the set-up time, the amount of work in process （WIP）, the capacity indicated by a Kanban, and the takt-time ratio. A novel method for optimizing performance on the premise of no stockouts is proposed. Empirical results show that the amount of WIP is reduced remarkably after optimization.

考虑物料配送的混流装配线协同优化

针对混流装配线的平衡排序和物料配送的协同问题,提出了一种装配线同步的协同配送策略,在此基础上以平衡工位负载、最小化最大完工时间和减少物料配送成本为目标,建立了考虑物料配送的协同优化模型。利用改进的非支配排序遗传算法(INSGA)对模型进行求解,基于模型的三层编码设计保证了解集的合法性和完整性。提出的多层编码交叉算子和变异算子,在提高种群多样性的同时也加快了算法的收敛速度。为避免算法陷入局部最优,重新设计了拥挤度距离,并改进了精英选择策略。此外,基于竞争机制,设计了一种舒适度竞争算子,便于在Pareto最优解集中确定具体方案。最后,通过标准算例和具体案例的求解,证明了改进算法的优越性和模型的有效性,案例的优化结果表示,装配线的完工时间缩短了近20%,物料配送成本节省了30%以上,取得了较好的优化效果。

汽车工厂Painted Body Storage的仿真研究

文章是针对于汽车厂PaintedBodyStorage(PBS)的模拟仿真研究。汽车厂装配车间是典型的混流生产方式的车间。在喷漆车间,为了减少颜色频繁变换带来的喷头清洗费用,需要尽量将相同颜色的车辆进行组合。PBS是位于油漆车间和装配车车间之间调配空间,控制投入到装配线的车辆顺序,装配线中要尽量要分散相同配置的车辆。满足车辆的接近禁止和均匀化等约束,一旦不能满足,就需要补助人员的投入。PBS的目标是尽量满足这两种约束条件。文章介绍了汽车工厂的PBS研究。提出了装配线的车辆顺序进场、出场的优化算法,通过模拟仿真证明了该算法的优越性,进行了运营因素中的灵敏度分析。

基于复合规则和强化学习的混流装配线调度方法

针对混流装配线的平衡与排序问题,提出了一种基于复合规则和强化学习的智能调度方法。根据数学模型,设计了平衡规则库与排序规则库,提出了规则加权组合的近端策略优化(PPO)算法,并利用具有Actor-Critic训练流程和优先经验回放机制的强化学习过程,实现了复合规则权值参数的调控优化,生成了平衡与排序方案。所提方法与PPO+单一规则算法、复合规则和遗传算法的对比实验验证了所提方法的有效性。

多约束混流线平衡与排序优化问题研究

针对G企业混流装配线设计不合理导致工位间负载不均衡及产品堆积现象,考虑工序与工位间匹配关系,节拍时间、工序优先关系等多约束,以减少工位数量、均衡工位间负载、减少产品等待时间为目标,建立混流装配线设计与产品排序问题的多目标混合整数规划模型,分别设计了分层算法和混合启发式算法;利用小规模算例验证了分层次算法的精确性,利用大规模算例验证了混合启发式算法高效性。基于本文算法对G企业装配线进行优化后,其工位间负载极差由15.6 min减少到0.1 min,优化后投产顺序对应的节拍时间降低了6 min。

改进鲸鱼优化算法及其在混流U型装配线平衡问题的应用

为提高鲸鱼优化算法的寻优效果和求解稳定性,通过改进算法的种群初始化、局部搜索、越界处理方法和收敛因子,提出了一种改进的鲸鱼优化算法。采用基于切割法的均匀设计种群初始化方法,改善算法初始种群的均匀性;通过引入基于NEWUOA算法的局部搜索算子,提高算法的局部搜索能力;在算法的越界处理方面,提出一种基于环形区间和随机波动的方式,降低算法陷入局部最优的可能;引入了非线性收敛因子和自适用权重,均衡算法的局部和全局搜索,并进一步增强搜索的精细度。通过7个单模态、多模态以及固定维度的基准测试函数进行了数值仿真实验,验证了改进的鲸鱼优化算法相较于遗传算法、鲸鱼优化算法以及其他改进的鲸鱼优化算法,在寻优效果和求解稳定性方面具有优越性。针对混流U型装配线平衡问题,考虑最小化装配线的节拍时间,将改进的鲸鱼优化算法用于问题求解;在解码阶段,设计一种基于阈值的解码方法,优化工序的分配过程;最后计算了21个混流装配线算例,结果表明,改进的鲸鱼优化算法在20个算例中求得了更优解,相较于其他算法,节拍时间平均降低3.02%。

基于物料超市的混流装配线准时化物料配送调度

针对混流装配线物料配送优化问题,结合新型的物料超市配送模式,提出了区别于传统周期性配送的准时化配送方法。首先,根据问题描述和相关假设构建了以线边库存最小化为目标的混合整数规划模型。然后,考虑到实际生产的连续性和复杂性,为求解大规模问题,设计出适用于问题的启发式算法,并与遗传算法进行比较,启发式算法能在更短时间内获得满意解。最后,在不同牵引拖车容量下,通过实例分析,与广泛应用的周期性配送进行比较,算例求解结果表明,准时化配送模型能够获得更低的库存水平,减少工作站的空间占用。

考虑工时不确定性的混流U型装配线平衡的分布鲁棒优化方法

在以人工装配为主的流水线生产中,工时的不确定性是影响生产节拍的重要因素。考虑到随机优化要求精确的概率分布信息和较高的鲁棒优化保守性,本文针对工时不确定条件下混流U型装配线平衡问题,采用以经验分布为中心、Wasserstein距离为半径的模糊集对工时的不确定性进行描述,并以最小化生产节拍为优化目标,建立装配线平衡问题的分布鲁棒优化模型。为了降低模型的复杂性,利用强对偶理论将模型转换为易于求解的形式;为保证解的鲁棒性,设计了一种鲁棒性指标并将其作为模型的约束条件。针对上述模型,通过设计一种基于区间数的解码方式,并引入自适应交叉和变异概率,给出了一种改进的遗传算法。最后通过标准算例和断路器抽架生产实例进行了数值仿真实验,结果表明相较于随机优化和鲁棒优化方法,所建立模型在降低结果保守性的同时保持较高的鲁棒性,并且针对问题所提出的改进遗传算法具有良好的寻优能力。

改进遗传算法支持下汽车中控混流装配生产线优化数学模型构建

针对当下对混流装配生产线现存方法不足的问题,实验提出一种结合GA算法和模拟退火算法的IGA模型,用以求解汽车混流装配线的平衡问题。模型对IGA所有步骤进行了充分的解释说明,并引入模拟退火算法中的采样规则来选择操作。最终实验结果显示,优化前后,生产线平衡率提高了4.77%,生产线负荷均衡提高了3.56,平衡损耗系数降低了0.047,生产节拍降低了6.21 s。该模型能够实现装配工艺的最优分配策略,同时提升整体生产线的工作效率。

改进智能水滴算法求解多目标混流装配线平衡问题

针对第Ⅰ类混流装配线平衡问题,为了更大限度地提高装配线效率,设计一种基于改进智能水滴算法的多目标混流装配线模型.该模型通过在选择工位的同时考虑启发式因素与土壤值来提高装配效率,在处理产品切换引起的负荷变化的同时,能够很好地优化装配线的工位数、平滑指数和操作关联度.通过使用帕累托分层对传统的智能水滴算法进行修改,从而达到一个帕累托集以实现目标的优化.利用若干个装配线问题的标杆算例对提出的算法进行测试并进行指标分析,实验结果显示与其他算法相比,本研究算法计算出的结果更优,非支配解比例、解集分布性等指标更好.

用于多目标双层优化的嵌套遗传算法——以混装线规划为例

多目标优化问题(Multi-objective Optimization Problem,MOP)是针对两个或多个冲突目标的决策问题。传统的解决方案的顺序优化方法只能找到单个问题的最优解,很难获得整体最优解;而归一化方法以加权和的形式简化了复合冲突目标函数,但权重是高度主观的,且会牺牲一些目标,因而都难以得到理想的结果。MOP的目标是协调和合并所有目标。为此提出了一个多目标双层优化模型。通过两级层次优化机制建立了两个相互关联的目标之间的协调模型。针对双层规划模型,提出了一种双层嵌套遗传算法及其相应的编码策略。以一个混合模型焊装生产线的平衡和排序问题为例,将所提算法和传统归一化方法、顺序优化方法进行对比,证明了所提出的多目标双层优化模型的可行性和有效性。