Practical Manufacturing Systems Research Articles

A Learning-Based Multipopulation Evolutionary Optimization for Flexible Job Shop Scheduling Problem With Finite Transportation Resources

In many practical manufacturing systems, transportation equipment such as automated guided vehicles is widely adopted to transfer jobs and realize the collaboration of different machines, but is often ignored in current researches. In this paper, we address the flexible job shop scheduling problem with finite transportation resources (FJSP-T). Considering the difficulties caused by the introduction of transportation and the NP-hard nature, the evolutionary algorithm is adopted as a solution approach. To this end, a learning-based multi-population evolutionary optimization (LMEO) is proposed to deal with the FJSP-T. First, the multi-population strategy is introduced and a cooperation-based initialization is designed by combining several heuristics to guarantee the quality and diversity of the initial population. Second, a reinforcement learning-based mating selection is proposed to realize the cooperation of different sub-populations by selecting appropriate individuals for evolutionary search. Then, a specific local search inspired by the problem properties is designed to enhance the exploitation capability of the LMEO. Moreover, a statistical learning-based replacement is designed to maintain the quality and diversity of the population. Extensive experiments are conducted to test the performances of the LMEO. The statistical comparison shows that the LMEO is superior to the state-of-the-art algorithms in solving the FJSP-T in terms of solution quality and robustness.

Hybrid flow-shop manufacturing network reliability optimization using genetic algorithm and absorbing Markov chain

In this study, we adopt a strategy of machine configuration to construct a reliable hybrid flow-shop manufacturing system (HFSMS) with reworking and scrapping actions, in which the reworking and scrapping actions are due to the yield rates of the configured machines. The machine configuration determines the machine suppliers and the number of machines for each production stage in the HFSMS, which may affect the stability of the HFSMS. The machines configured to a production stage in parallel are provided by the same supplier. Accordingly, each production stage has multiple states, following a probability distribution. The HFSMS is a typical stochastic-flow network under machine configuration, and network reliability indicates the probability that order demand d can be fulfilled by the HFSMS and used as a decision reference. Our study integrates genetic algorithm (GA) and absorbing Markov chain (AMC) to solve the reliability-oriented machine configuration problem of HFSMS subject to a configuration budget, where the GA is utilized to find the optimal machine configuration with maximum network reliability and the AMC model constructed based on the yield rates is used for network reliability evaluation. For validating the applicability and computational efficiency of the proposed AMC and GA-based approach, a simple manufacturing system and two practical manufacturing systems are used to compare the proposed approach with four popular meta-heuristic algorithms. The experimental results show that the proposed approach can find the exact optimal solution and has better computational efficiency than the other four meta-heuristic algorithms.

Learning-Based Grey Wolf Optimizer for Stochastic Flexible Job Shop Scheduling

This work considers a stochastic flexible job shop scheduling with limited extra resources and machine-dependent setup time in a semiconductor manufacturing environment, which is an NP-hard problem. In order to obtain its reliable and high-performance schedule in a reasonable time, a learning-based grey wolf optimizer is proposed. In it, an optimal computing budget allocation-based approach, which is designed for two scenarios from real manufacturing environments, is proposed to intelligently allocate computing budget and improve search efficiency. It extends the application area of optimal computing budget allocation by laying a theoretic foundation. Besides, to obtain proper control parameters iteratively, a reinforcement learning algorithm with a newly designed delay update strategy is used to build a parameter tuning scheme of a grey wolf optimizer. The scheme acts as a guide for balancing global and local search, thereby enhancing effectiveness of the proposed algorithm. The theoretic interpretation of the developed optimal computing budget allocation-based approach and the convergence analysis results of the proposed algorithm are presented. Various experiments with benchmarks and randomly generated cases are performed to compare it with several updated algorithms. The results shows its superiority over them. Note to Practitioners—Meta-heuristic are often deployed to solve semiconductor manufacturing scheduling problems. However, they face to two thorny issues when they face stochastic manufacturing environments. 1) their computational efficiency is quite low, thus requiring substantial improvement, since a stochastic optimization problem requires Monte Carlo sampling to estimate the actual objective function values in a precise manner; and 2) most of them are parameter-sensitive, and choosing their proper parameters is highly challenging in such environments. To address the first issue, we develop an optimal computing budget allocation-based method for deciding the optimal numbers of sampling times based on both prior knowledge and simulation results. To address the second one, we propose a reinforcement learning algorithm to self-adjust the parameters of our proposed method called Learning-based Grey Wolf Optimizer. In addition, we design a delay update strategy to enhance its robustness, and thus, a feasible and high-quality schedule can be founded in a short time for real-time scheduling problems. Theoretic proofs and experimental results show that the proposed method is effective and efficient. Consequently, it can be readily applicable to practical semiconductor manufacturing systems.

A distributed flexible job shop scheduling problem considering worker arrangement using an improved memetic algorithm

The classical distributed flexible job shop scheduling problem (DFJSP) mainly considers factory allocation, machine arrangement, job sequencing and transportation. To date, the relevant literature has not studied the DFJSPs with worker arrangement, which widely exists in practical manufacturing systems. In this paper, we investigate the DFJSP with worker arrangement (DFJSPW), where not only the factories, machines and operations, but the workers are considered simultaneously. A mixed-integer linear programming model is formulated for this problem. Correspondingly, an improved memetic algorithm (IMA) based on the structure of NSGA-II is proposed for the proposed DFJSPW whose objective is to minimize the makespan, maximum workload of machines and workload of workers simultaneously. In IMA, a simplified two-level encoding and four heuristic decoding methods are designed to encode and decode the individuals. A well-designed adaptive neighborhood search operator is developed to enhance the local search ability of IMA and speed its convergence. Fifty-eight benchmarks are constructed to evaluate the performance of our proposed IMA. Extensive experiments show that in most examples, IMA performs better than four well-known multi-objective algorithms, demonstrating the superiority of IMA in solving the DFJSPW.

Energy-efficient flexible flow shop scheduling with worker flexibility

The classical flexible flow shop scheduling problem (FFSP) only considers machine flexibility. Thus far, the relevant literature has not studied FFSPs with worker flexibility, which is widely seen in practical manufacturing systems. Worker flexibility may greatly affect production efficiency and productivity. Furthermore, with the increase of environmental pollution and energy consumption, manufacturers require innovative methods to improve energy efficiency. In this paper, we propose an energy-efficient FFSP with worker flexibility (EFFSPW), in which the flexibility of machines and workers as well as the processing time, energy consumption and worker cost related factors are considered simultaneously. A hybrid evolutionary algorithm (HEA) is then presented to solve the proposed EFFSPW, where some effective operators and a new variable neighborhood search approach are designed. Comprehensive experiments including 54 benchmark instances of the EFFSPW are carried out, and Taguchi analysis is used to determine the best combination of key parameters for the HEA. Experimental results show that the proposed HEA can obtain better solutions for most of these benchmark instances compared to two other well-known algorithms, demonstrating its superior performance in terms of both solution quality and computational efficiency.

A Three-Stage Approach to a Multirow Parallel Machine Layout Problem

Facility layout is vital to save operational cost and enhance production efficiency. Multirow layout is a common pattern in practical manufacturing environment. Although parallel machines are frequently implemented in practice to enhance productivity, there lacks any in-depth study on multirow layout problem with parallel machines. In this paper, its mathematical programming formulation is established to minimize material flow cost. A three-stage approach is proposed to solve it. First, a Monte Carlo heuristic is devised to optimize the sequence of machines on multiple rows. Second, a linear program is used to determine the optimal exact location of each machine. Finally, an exchange heuristic is adopted to reassign material flows among parallel machines in different machine groups. An iterative optimization strategy is suggested to execute the three stages repeatedly to improve the solution quality. This approach is applied to a number of problem instances and compared against others. The experimental results show that it is able to effectively solve this new problem and significantly decrease material flow cost. Note to Practitioners —Multirow layout is common in practical manufacturing systems in which parallel machines are often used to improve productivity, shorten production time, and guarantee some flexibility. This paper studies a multirow parallel machine layout problem that involves machine groups, each of which contains parallel machines. Solving it is to locate all machines at multiple parallel rows to minimize material flow cost. It is challenging because one needs to decompose material flows and determine exact locations of machines simultaneously. A three-stage approach is proposed to do so. It is applied to many problem instances. The results demonstrate that it works well for such layout problems with parallel machines.

A Robust Prevention Method for Automated Manufacturing Systems With Unreliable Resources Using Petri Nets

On the basis of the assumption that no resource failure occurs, a variety of deadlock control policies have been developed for automated manufacturing systems (AMSs). However, in practical manufacturing systems, the occurrence of resource failures is always inescapable. In case, if resources fail to work, the existing deadlock control strategies can no longer be applied to the changed system; therefore, redesigning a new strategy is necessary. Because of their powerful modeling capabilities, Petri nets are used to model the considered AMSs allowing multi-type and multi-quantity resource acquisition. This paper focuses on extending a deadlock prevention method to be applied to AMSs with unreliable resources. Strict minimal siphons are controlled by added control places (monitors) to ensure the system’s liveness. To prevent blocking issues caused by resource failures, we develop a set of shared resource constraints represented by a set of inequalities based on the minimal resource requirements of processes and the capacity of shared resources. Robust monitors are designed for them to limit the distribution of tokens in unreliable neighborhood places. Our objective is to control resource allocation such that those parts not necessarily requiring any failed resource can continue progressing smoothly even if some unreliable resources break down. Examples are given to elucidate our proposed method clearly.

Transient Analysis of Geometric Serial Lines With Perishable Intermediate Products

Perishable intermediate products are commonly seen in practical manufacturing systems, where the time that intermediate products spend in a buffer is limited by a certain number. Parts have to be scrapped if their maximum allowed time is exceeded. In order to reduce the scrap ratio and optimize the production operation in a timely manner, the transient behavior of the system needs to be investigated to evaluate and predict the system performance at any time instance. In this paper, we develop a Markov chain model to analyze the transient behaviors of a two-machine geometric production system accounting for scrap due to the perishability of intermediate products. The analytical formulas are derived to evaluate transient performance. Compared with simulation, it is shown that the proposed method could estimate the system's transient performance with high accuracy.

A Single Machine Capacitated Production Planning Problem Under Uncertainty: A Grey Linear Programming Approach

The production planning is an important problem in most of manufacturing systems in practice. Unlike many researches existing in literature, this problem encounters with great uncertainties in parameters and input data. In this paper, a single machine capacitated production planning problem is considered and a linear programming formulation is presented. The production costs are assumed to be uncertain parameters. To handle the uncertainties in the model, the grey systems theory is employed and the concept of grey numbers is incorporated into an optimization framework. In such systems, the uncertain parameters with unknown distributions can be handled by grey numbers. The grey linear programming (GLP) is a development of the classical linear programming which allows uncertainty to be effectively communicated into the optimization process. Finally, the uncertain problem is transformed into a GLP, and is solved by two linear deterministic sub-models.

Application of multi-fidelity simulation modelling to integrated circuit packaging

In semiconductor manufacturing, time-to-market is critical to maintain a competitive advantage through achieving customer satisfaction. Inefficient ways of utilising production resources will lead to a long cycle time. Therefore, machine allocation becomes an essential production decision in many practical manufacturing systems, especially for integrated circuit (IC) packaging. IC packaging is the process of encasing the finished die in a package in order to prevent corrosion and physical damage. Advanced IC packaging techniques add even more complexity into the production system, so reliable average cycle time of this complex system becomes difficult to obtain. We propose a new simulation optimisation framework with multi-fidelity models to study an IC packaging case of the machine allocation problem to pursue a minimum average cycle time. This framework consists of two methodologies: ordinal transformation (OT) and optimal sampling (OS). The OT first employs the low-fidelity model to fast observe all designs, and extracts insightful information from this model by transforming the original design space into an ordinal space. It follows that OS efficiently allocates the computing budget for searching the best design via high-fidelity simulations. An empirical study based on real data was conducted to validate the practical viability of the proposed framework.

An integrated control strategy to solve the disturbance decoupling problem for max-plus linear systems with applications to a high throughput screening system

This paper presents the new investigations on the disturbance decoupling problem (DDP) for the geometric control of max-plus linear systems. The classical DDP concept in the geometric control theory means that the controlled outputs will not be changed by any disturbances. In practical manufacturing systems, solving for the DDP would require further delays on the output parts than the existing delays caused by the system breakdown. The new proposed modified disturbance decoupling problem (MDDP) in this paper ensures that the controlled output signals will not be delayed more than the existing delays caused by the disturbances in order to achieve the just-in-time optimal control. Furthermore, this paper presents the integration of output feedback and open-loop control strategies to solve for the MDDP, as well as for the DDP. If these controls can only solve for the MDDP, but not for the DDP, an evaluation principle is established to compare the distance between two output signals generated by controls solving for the MDDP and DDP, respectively. This distance can be interpreted as the number of tokens or firings that are needed in order for the controls to solve for the DDP. Moreover, another alternative approach is finding a new disturbance mapping in order to guarantee the solvability of the DDP by the same optimal control for the MDDP. The main results of this paper are illustrated by using a timed event graph model of a high throughput screening system in drug discovery.

Resource Allocation and Job Dispatching for Unreliable Flexible Flow Shop Manufacturing System

Resource allocation, product batching and production scheduling are three different problems in manufacturing systems of different structures such as flexible flow shop manufacturing systems. These problems are usually dealt with independently for a certain objective function related to production efficiency and effectiveness. Handling all of them in an integrated manner is a challenge facing many manufacturing systems in practice and that challenge increases for highly complicated and stochastic systems. Random arrival of products, machine setup time requirements, unexpected machine breakdowns, and multiple conflicting objective functions are some of the common complications in such systems. This research attempts to study the integrated problem under the mentioned complications with various objective functions. The decisions parameters are the batch size, the number of machines at each workstation, and the dispatching policy. Discrete event simulation is used as an optimization tools. The system is modeled using the ARENA software and different scenarios are tested for optimum parameter selection under different conditions.

Resource Allocation and Job Dispatching for Unreliable Flexible Flow Shop Manufacturing System

Resource allocation, product batching and production scheduling are three different problems in manufacturing systems of different structures such as flexible flow shop manufacturing systems. These problems are usually dealt with independently for a certain objective function related to production efficiency and effectiveness. Handling all of them in an integrated manner is a challenge facing many manufacturing systems in practice and that challenge increases for highly complicated and stochastic systems. Random arrival of products, machine setup time requirements, unexpected machine breakdowns, and multiple conflicting objective functions are some of the common complications in such systems. This research attempts to study the integrated problem under the mentioned complications with various objective functions. The decisions parameters are the batch size, the number of machines at each workstation, and the dispatching policy. Discrete event simulation is used as an optimization tools. The system is modeled using the ARENA software and different scenarios are tested for optimum parameter selection under different conditions.

An iterative method for synthesizing non-blocking supervisors for a class of generalized Petri nets using mathematical programming

This paper presents a novel and computational deadlock prevention policy for a class of generalized Petri nets, namely G-systems, which allows multiple resource acquisitions and flexible routings with machining, assembly and disassembly operations. In this research, a mixed integer programming (MIP)-based deadlock detection technique is used to find an insufficiently marked minimal siphon from a maximal deadly marked siphon for generalized Petri nets. In addition, two-stage control method is employed for deadlock prevention in Petri net model. Such proposed method is an iterative approach consisting of two stages. The first one is called siphons control, which adds a control place to the original net for each insufficiently marked minimal siphon. The objective is to prevent minimal siphons from being insufficiently marked. The second one, called control-induced siphons control, is to add a control place to the augmented net with its output arcs connecting to source transitions, which assures that there is no new insufficiently marked siphon generated due to the addition of the monitors. Compared with the existing approaches, the proposed deadlock prevention policy can usually lead to a non-blocking supervisor with more permissive behavior and high computational efficiency for a sizeable plant model due to avoiding complete siphon enumeration. Finally, a practical flexible manufacturing system (FMS) example is utilized to illustrate the proposed method.

Queueing models for a single machine subject to multiple types of interruptions

Queueing models are commonly applied to quantify the performance of production systems. Prior research has usually focused on deriving queueing models for a specific type of interruptions. However, machines generally suffer multiple types of interruptions in practical manufacturing systems. To satisfy this need, an integrated model is proposed, in which multiple types of interruptions commonly seen on the shop floor are considered.

Overall Design of the Forming System for Squeeze Casting

As a forming process, squeeze casting is completed in the corresponding manufacturing system. The paper deals with the overall design of the manufacturing system-the forming system for squeeze casting. To construct a practical manufacturing system for this technology, the forming control process of squeeze casting is analyzed, where the key factors are identified to determine the production quality. Then the design principles for the forming system for squeeze casting are built and its functional requirements are attained. And an overall structure for this forming system is presented including the die design system, the design and management system of process parameters and the squeeze casting equipment system. This structure decreases the difficulties and the cost of building the physical system.

Practical Extensions to Cycle Time Approximations for the $G/G/m$-Queue With Applications

Approximate closed form expressions for the mean cycle time in a G/G/m-queue often serve as practical and intuitive alternatives to more exact but less tractable analyses. However, the G/G/m-queue model may not fully address issues that arise in practical manufacturing systems. Such issues include tools with production parallelism, tools that are idle with work in process, travel to the queue, and the tendency of lots to defect from a failed server and return to the queue even after they have entered production. In this paper, we extend popular approximate mean cycle time formulae to address these practical manufacturing issues. Employing automated data extraction algorithms embedded in software, we test the approximations using parameters gleaned from production tool groups in IBM's 200 mm semiconductor wafer fabricator.

Design of Liveness-Enforcing Supervisors for Flexible Manufacturing Systems Using Petri Nets

This correspondence develops a methodology to synthesize supervisors for a special class of Petri nets that can well model many flexible manufacturing systems. In our previous work, siphons in a Petri net are divided into elementary and dependent ones on the condition that all emptiable minimal siphons are computed. In this research, a mixed integer programming (MlP)-based deadlock detection technique is used to find some, in general not all, minimal siphons in a plant model without complete siphon enumeration. For each siphon found, depending on its noncontrollability, a monitor is added such that it is invariant-controlled. Our siphon control method guarantees that no emptiable control-induced siphon is generated due to the addition of the monitors. The siphon control process proceeds iteratively until there is no unmarked siphon in the supervisor of a plant model. Compared with the existing approaches, the novel deadlock prevention policy can usually lead to a structurally simple liveness-enforcing supervisor by adding only a small number of monitors and arcs. More importantly, complete siphon enumeration is avoided. A practical flexible manufacturing system (FMS) example is utilized to illustrate the proposed methods.

Optimization of manufacturing processes by distributed simulation

The paper describes a simulation-based optimization system for manufacturing processes. The principle suitability of simple Threshold Accepting and Great Deluge Algorithms for optimizing practical manufacturing systems is shown, as well as the possibility of parallelizing of these algorithms. A special IP Multicast architecture is introduced which allows the simulation model to be extended to multiple clients and so the efficiency of the optimization algorithms to be increased by concurrent simulation. Each client can log on or log out at an optimization server at any time so that it should be possible to use the idle computer capacity of a single production department or a whole company. The optimization server is implemented as an experimental option in the simulation system simcron MODELLER®, which is particularly used in electronics and semiconductor manufacturing.

Improved performance evaluation of tool condition identification by manufacturing loss consideration

When neural networks are used to identify tool states in machining processes, the main interest is often the recognition ability. It is usually believed that a higher classification rate from pattern recognition can improve the accuracy and reliability of tool condition monitoring, thereby reducing the manufacturing loss. Nevertheless, the two objectives are not identical in most practical manufacturing systems. The aim is to address this issue and propose a new performance evaluation function so that the recognition ability of tool condition monitoring can be evaluated more reasonably. On this basis, two kinds of manufacturing loss due to misclassification are analysed: the over-prediction caused by misclassifying the worn tool condition; and the under-prediction caused by misclassifying the fresh tool condition. By using both to calculate corresponding weights in the performance evaluation function, the potential manufacturing loss is introduced to evaluate the recognition performance of tool condition monitoring. Based on this performance evaluation function, a modified support vector machine approach with two regularization parameters is employed to learn the information of every tool state. In this support vector machine design, the effective feature set extracted from acoustic emission signals is used as inputs, and a five-fold cross-validation is used to tune the parameters. The experimental results show that the proposed method can reliably identify tool flank wear and reduce the overdue prediction of worn tool conditions and its relative loss. Experimental results show that this approach may effectively identify tool state over a range of cutting conditions and reduce the manufacturing loss in the practical industry process.