A flow matrix-based heuristic algorithm for cell formation and layout design in cellular manufacturing system

Cellular manufacturing is a successful application of group technology (GT) concepts. The aim of a cellular manufacturing system (CMS) is to identify similar manufacturing processes and features where machines are grouped into machine cells based on their contributions to the production process. In the last three decades of research in cell formation, researchers have mainly used zero-one machine component incidence matrix as the input data for the problem. However, recently efforts have been made to use other data structures, such as interval data and ordinal data (consisting of sequence of processing). Sequence data provide valuable information about the flow patterns of various jobs in a manufacturing system. This study develops a heuristic algorithm based on flow matrix for cell formation and layout design in a simultaneous fashion using sequence data. The numerical results of the algorithm on the available problems in the literature indicate the usefulness of the algorithm regarding to performance indices.     

Concurrent cell formation and layout design using scatter search

Two key decisions in designing cellular manufacturing systems are cell formation and layout design problems. In the cell formation problem, machine groups and part families are determined while in the facility layout problem the location of each machine in each cell (intra-cell layout) and the location of each cell (inter-cell layout) are decided. Owing to the fact that there are interactions between two problems, cell formation and layout design problem must be tackled concurrently to design a productive manufacturing system. In this research, two problems are investigated concurrently. Some important and realistic factors such as inter-cell layout, intra-cell layout, operations sequence, part demands, batch size, number of cells, cell size, and variable process routings are incorporated in the problem. The problem is formulated as a mathematical model. Three different methods are described to solve the problem: multi-objective scatter search (MOSS), non-dominated genetic algorithm (NSGA-II), and the ε-constraint method. The methods are employed to solve nine problems generated and adopted from the literature. Sensitivity analysis is accomplished on the parameters of the problem to investigate the effects of them on objective function values. The results show that the proposed MOSS algorithm performs better than NSGA-II and produces better solutions in comparison to multi-stage approaches.     

A comprehensive approach to facility layout design and cell formation

Group technology (GT) and a facility layout problem share a common factor to optimize, which is the “Intercell Flow or Material Handling Flow”. Most group technology-based approaches use binary part-machine incidence matrix for cell formation. Impacts of flow volume analysis between machine pairs have not been analyzed in the design of cellular manufacturing systems. It is important to analyze the placement of machine cells for reduced material handling costs at the grouping stage itself. This paper develops a method to integrate GT philosophy with a facility layout problem using “From–Between” relationship chart as a tool. A new optimization technique with a “from–to chart” as the primary input, is used to determine the cell formation. A modified grouping efficiency measure is used to determine the efficiency of grouping. A genetic algorithm-based approach is used for machine cell placement within a layout matrix. Cells formed using the proposed approach show reduced intercell flows, high grouping efficiencies and reduced material handling costs when cells are placed in a layout grid compared to existing cell formation approaches.     

Implementing lean manufacturing with cellular layout: a case study

Lean manufacturing is an applied methodology of scientific, objective techniques that cause work tasks in a process to be performed with a minimum of non-value adding activities resulting in greatly reduced wait time, queue time, move time, administrative time, and other delays. In a cellular manufacturing system (CMS), machines are grouped into several cells, where each cell is dedicated to a particular part family and the objective is to maximize cell independence. CMS helps in reducing the material handling, work-in-process, setup time, and manufacturing lead time and improve productivity, operation control, etc. The facility layout used during lean implementation can be either be a line layout or in the form of cells. After grouping parts in to various part families, machine cells can be formed to produce those parts well inside the cells. As some of the lean manufacturing concepts are different from that of cellular manufacturing, e.g., establishment of Takt time, Takt-based resource balancing, etc., some new cell design methodology is required to be explored that is compatible with lean manufacturing. The rate at which work progresses through the factory is called flow rate or Takt. In the present work, a design methodology for cellular layout is proposed for implementing lean concepts and is exemplified in a manufacturing industry dealing with ammunition components for defense applications. Based on Takt time for various parts, the production flow among cells was optimized thus minimizing several non-value added activities/times such as bottlenecking time, waiting time, material handling time, etc. This case study can be useful in developing a more generic approach to design cellular layouts in lean environment.     

Mathematical modelling of multi-objective job shop scheduling with dependent setups and re-entrant operations

For the last thirty years, in job shop scheduling, the setup times of operations have been either ignored or combined with their corresponding processing times, independent setups, to simplify the work of model construction. However, in practice, the setup times of an operation are usually considered be sequence-dependent. Moreover, the performance of a scheduling system is not evaluated to satisfy a single objective, but to obtain a trade-off solution regarding multiple objectives. Therefore, this paper investigates job shop scheduling problems with re-entrant operations where the setup times, which vary according to the preceding operation which is processed on the same machine, and which can be separated from their corresponding processing times, cannot be omitted. Three practical performance measures – minimum total job flow time, minimum total job tardiness and minimum machine idle time – are considered. An integer programming model is first developed to optimise each single objective and an acceptable trade-off schedule, which makes use of a multiple-decision-making technique, the global criterion method, is obtained by evaluating three objectives simultaneously .     

Solving cell formation problem in cellular manufacturing using ant-colony-based optimization

Grouping the machines and parts in a cellular manufacturing system based on similarities is known as the cell formation problem. It has been shown that cell formation problem is a NP-hard problem. In this paper, the ant colony optimization (ACO) method is used as an evolutionary approach to solve the cell formation problem. This model uses a P?=?[P _ij ] _{(C)?×?(M?+?P)} pheromone matrix in which C, M, and P are the number of cells, machines, and parts, respectively. In order to represent the sequence of operations, the machine–part incidence matrix entries are considered as positive integers. Performance of the proposed algorithm is tested on some benchmark problems existing in the literature to show the applicability and effectiveness of the proposed model. Comparison of the solutions obtained by the proposed algorithm with those reported in the literature indicates that application of the proposed algorithm has resulted in 5.73% improvement in the total number of intercellular movements and voids on average.     

敏捷制造车间布局优化的启发式算法

为了有效地减少产品在敏捷制造车间中的非加工费用,根据车间中的生产工艺流程对设备进行优化布局,提出了一个处理敏捷制造车间机器布局的启发式算法,算法综合考虑机器间零件流的流量、零件的传输因素、机器的空间布局限制等,以机器间的传输费用作为评价分值,以总分值最小为优化目标,通过收养、合并、交叉、移位操作,依次将各机器安排在一个3×3的网格布局中。该算法提供了一个灵活的、近似最优的车间设备布局。最后,演示了一个虚拟实例。     

An analytical-iterative clustering algorithm for cell formation in cellular manufacturing systems with ordinal-level and ratio-level data

In this paper, the problem of clustering machines into cells and components into part-families with the consideration of ratio-level and ordinal-level data is dealt with. The ratio-level data is characterized by the use of workload information obtained both from per-unitprocess times and production quantity of components, and from machine capacity. In the case of ordinal-level data, we consider the sequence of operations for every component. These data sets are used in place of conventional binary data for arriving at clusters of cells and part-families. We propose a new approach to cell formation by viewing machines, and subsequently components, as 'points' in multi-dimensional space, with their coordinates defined by the corresponding elements in a Machine-Component Incidence Matrix (MCIM). An iterative algorithm that improves upon the seed solution is developed. The seed solution is obtained by formulating the given clustering problem as a Traveling Salesman Problem (TSP). The solutions yielded by the proposed clustering algorithm are found to be good and comparable to those reported in the literature.     

Designing integrated cellular manufacturing systems with scheduling considering stochastic processing time

This paper addresses a new mathematical model for cellular manufacturing problem integrated with group scheduling in an uncertain space. This model optimizes cell formation and scheduling decisions, concurrently. It is assumed that processing time of parts on machines is stochastic and described by discrete scenarios enhances application of real assumptions in analytical process. This model aims to minimize total expected cost consisting maximum tardiness cost among all parts, cost of subcontracting for exceptional elements and the cost of resource underutilization. Scheduling problem in a cellular manufacturing environment is treated as group scheduling problem, which assumes that all parts in a part family are processed in the same cell and no inter-cellular transfer is needed. Finally, the nonlinear model will be transformed to a linear form in order to solve it for optimality. To solve such a stochastic model, an efficient hybrid method based on new combination of genetic algorithm (GA), simulated annealing (SA) algorithm, and an optimization rule will be proposed where SA and optimization rule are subordinate parts of GA under a self-learning rule criterion. Also, performance and robustness of the algorithm will be verified through some test problems against branch and bound and a heuristic procedure.     

Genetic cell formation using ratio level data in cellular manufacturing systems

Manufacturing cell formation is a useful strategy in batch type production industries for enhancing productivity and flexibility. The basic idea rests on grouping the parts into part families and the machines into machine cells. Most of the literature used zero-one incidence matrix representing the part visiting a particular machine as one and zero otherwise. The output is generated in the form of block diagonal structure where each block represents a machine cell and a part family. In such models real life production factors such as operation time and sequence of operations are not accounted for. In this paper, the operational time of the parts required for processing in the machines is considered. It is attempted to develop an algorithm using genetic algorithm (GA) with a combined objective of minimizing the total cell load variation and the exceptional elements. The results are compared with the solutions obtained from K-means clustering and C-linkage clustering algorithms.     

An intelligent operations scheduling system in a job shop

Scheduling jobs effectively under the consideration of actual loads on machines is one of the most complicated tasks in production control. The conventional scheduling methods fail because of the complexity of the tasks. To deal with the complexity, knowledge-based approaches to job shop scheduling have been attempted. This paper presents an interactive scheduling expert system, IOSS (Intelligent Operations Scheduling System), which performs both predictive and reactive scheduling. IOSS combines the knowledge-based scheduling method with the interactive scheduling method to generate a feasible schedule and to revise the existing schedule. It is based on opportunistic and interactive repair based problem solving within a blackboard architecture. To handle conflicting events, heuristics are applied from the order point of view. Flexible reaction management is possible while keeping the changes in the generated schedule to a minimum by adjusting the schedule for tardy operations or changes in job shop status. The effectiveness of the proposed concept is demonstrated by applying the developed system to an example case.     

A tradeoff analysis between process plan selection and cell formation in cellular manufacturing

In this paper, an algorithm to evaluate the tradeoff between conflicting objectives in process plan selection and cell formation is developed. Consideration of the minimisation of intercell material movement in cellular manufacturing is necessary but not in itself sufficient to produce a system for which the total work content is minimised. Solving the process plan selection and the cell formation problem for all possible alternative process plans is a time-consuming task, and therefore not economically justifiable. The algorithm is illustrated through the use of a sample problem that shows how it is possible to create a cell using the algorithm presented in this paper.     

基于协同粒子群算法的单元集成布局方法

为避免单元间布局和单元内设施布局分开孤立研究所导致的问题解空间损失,建立了多目标集成优化模型,对布局中涉及的单元间布局、单元内设施布置、设施摆放方向三个问题同时进行描述,扩大问题解的搜索空间。针对模型的复杂性和传统粒子群算法容易陷入局部最优的弱点,设计了基于结构化编码的多粒子群协同进化算法,在多个粒子群协同进化的同时,对各子群构建基因库,并对基因库粒子进行遗传操作,提高了算法的寻优能力。通过求解单元布局问题实例,验证了模型和算法的有效性。     

基于差分元胞多目标遗传算法的车间布局优化

以物料搬运费用最小和车间设备占地面积利用率最大为目标,建立了车间设备布局多目标优化设计模型.针对常用多目标算法不能很好求解该模型的问题,提出一种差分元胞多目标遗传算法.该算法在经典元胞多目标遗传算法的基础上引入差分演化策略,从而集成了元胞算法多样性好和差分演化策略在解决复杂问题时收敛性强、覆盖范围广的特点.分别运用该算法、经典元胞多目标遗传算法和NSGAII对测试函数及车间设备布局模型进行计算,通过数据和性能比较分析表明,针对多约束、多变量、非线性的模型,新算法具有良好的收敛性、分布性和扩展性,能有效解决相关生产实践问题.     

Integrating lotsizing, loading, and scheduling decisions in flexible flow shops

Presenting an integrated lotsizing, loading, and scheduling model for the capacitated flexible flow shops with sequence-dependent setups is the main contribution of this paper. An exact formulation of the problem is provided as a mixed integer program. To solve this problem, mixed integer programming-based heuristics based on iterative procedures are provided. To test the accuracy of heuristics, two lower bounds are developed and compared against the optimal solution. The trade-offs between solution quality and computational time of heuristics are also provided.     

A hybrid metaheuristic for concurrent layout and scheduling problem in a job shop environment

In order to obtain efficiency and flexibility, assignment of machines’ layout and determination of jobs’ schedule on each machine are among the most important decisions. These decisions are interrelated and may impact each other but they are often treated separately or as a sequential decision in prior researches. In this paper, we propose a new approach to concurrently make the layout and scheduling decisions in a job shop environment. In other words, we consider an extension of the well-known job shop scheduling problem with transportation delay in which in addition to decisions made in the classic problem, the locations of machines have to be selected among possible sites. The only goal of the problem is the minimization of the makespan. A hybrid metaheuristic approach based on the scatter search algorithm is developed to tackle this problem. Using 43 randomly generated benchmark instances, the performance of the scatter search and its components are evaluated. We also applied our procedure to the classic job shop scheduling problems. Computational results show that our procedure is efficient.