A hybrid electromagnetism-like algorithm for two-stage assembly flow shop scheduling problem

This paper presents a study on the two-stage assembly flow shop scheduling problem for minimising the weighed sum of maximum makespan, earliness and lateness. There are m machines at the first stage, each of which produces a component of a job. A single machine at the second stage assembles the m components together to complete the job. A novel model for solving the scheduling problem is built to optimise the maximum makespan, earliness and lateness simultaneously. Two optimal operation sequences of jobs are determined and verified. As the problem is known to be NP-hard, a hybrid variable neighbourhood search – electromagnetism-like mechanism (VNS-EM) algorithm is proposed for its handling. To search beyond local optima for a global one, VNS algorithm is embedded in each iteration of EM, whereby the fine neighbourhood search of optimum individuals can be realised and the solution is thus optimised. Simulation results show that the proposed hybrid VNS-EM algorithm outperforms the EM and VNS algorithms in both average value and standard deviation.     

Scheduling a batch processing machine with non-identical job sizes: a clustering perspective

Batch processing machines that process a group of jobs simultaneously are often encountered in semiconductor manufacturing and metal heat treatment. This paper considered the problem of scheduling a batch processing machine from a clustering perspective. We first demonstrated that minimising makespan on a single batching machine with non-identical job sizes can be regarded as a special clustering problem, providing a novel insight into scheduling with batching. The definition of WRB (waste ratio of batch) was then presented, and the objective function of minimising makespan was transformed into minimising weighted WRB so as to define the distance measure between batches in a more understandable way. The equivalence of the two objective functions was also proved. In addition, a clustering algorithm CACB (constrained agglomerative clustering of batches) was proposed based on the definition of WRB. To test the effectiveness of the proposed algorithm, the results obtained from CACB were compared with those from the previous methods, including BFLPT (best-fit longest processing time) heuristic and GA (genetic algorithm). CACB outperforms BFLPT and GA especially for large-scale problems.     

Minimizing the makespan for the two-machine scheduling problem with a single server: Two algorithms for very large instances

This article considers the problem of scheduling a given set of n jobs on two identical parallel machines with a single server. Each job must be processed on one of the machines. Before processing, the server has to set up the relevant machine. The objective is to minimize the makespan. For this unary NP-hard problem, two fast constructive algorithms with a complexity of O(n²) are presented. The performance of these algorithms is evaluated for instances with up to 10,000 jobs. Computational results indicate that the algorithms have an excellent performance for very large instances so that the obtained objective function values are very close to a lower bound, and in many cases even an optimal solution is achieved. Superiority over all existing algorithms is obtained by sequencing the jobs on the two machines so that the machine idle time and the server waiting time are minimized. In doing so, the characteristics of an optimal solution resulting from its relevant lower bound are taken into account.     

A priority scheduling approach for flexible job shops with multiple process plans

This paper considers the job shop scheduling problem with alternative operations and machines, called the flexible job shop scheduling problem. As an extension of previous studies, operation and routing flexibilities are considered at the same time in the form of multiple process plans, i.e. each job can be processed through alternative operations, each of which can be processed on alternative machines. The main decisions are: (a) selecting operation/machine pair; and (b) sequencing the jobs assigned to each machine. Since the problem is highly complicated, we suggest a practical priority scheduling approach in which the two decisions are done at the same time using a combination of operation/machine selection and job sequencing rules. The performance measures used are minimising makespan, total flow time, mean tardiness, the number of tardy jobs, and the maximum tardiness. To compare the performances of various rule combinations, simulation experiments were done on the data for hybrid systems with an advanced reconfigurable manufacturing system and a conventional legacy system, and the results are reported.     

A bi-objective parallel machine problem with eligibility,release dates and delivery times of the jobs

The scheduling of parallel machines is a well-known problem in many companies. Nevertheless, not always all the jobs can be manufactured in any machine and the eligibility appears. Based on a real-life problem, we present a model which has m parallel machines with different level of quality from the highest level for the first machine till the lowest level for the last machine. The set of jobs to be scheduled on these m parallel machines are also distributed among these m levels: one job from a level can be manufactured in a machine of the same or higher level but a penalty, depending on the level, appears when a job is manufactured in a machine different from the highest level i.e. different from the first machine. Besides, there are release dates and delivery times associated to each job. The tackled problem is bi-objective with the criteria: minimisation of the final date – i.e. the maximum for all the jobs of their completion time plus the delivery time – and the minimisation of the total penalty generated by the jobs. In a first step, we analyse the sub-problem of minimisation of the final date on a single machine for jobs with release dates and delivery times. Four heuristics and an improvement algorithm are proposed and compared on didactic examples and on a large set of instances. In a second step an algorithm is proposed to approximate the set of efficient solutions and the Pareto front of the bi-objective problem. This algorithm contains two phases: the first is a depth search phase and the second is a backtracking phase. The procedure is illustrated in detail on an instance with 20 jobs and 3 machines. Then extensive numerical experiments are realised on two different sets of instances, with 20, 30 and 50 jobs, 3 or 4 machines and various values of penalties. Except for the case of 50 jobs, the results are compared with the exact Pareto front.     

On bilevel machine scheduling problems

Bilevel scheduling problems constitute a hardly studied area of scheduling theory. In this paper, we summarise the basic concepts of bilevel optimisation, and discuss two problem classes for which we establish various complexity and algorithmic results. The first one is the bilevel total weighted completion time problem in which the leader assigns the jobs to parallel machines and the follower sequences the jobs assigned to each machine. Both the leader and the follower aims to minimise the total weighted completion time objective, but with different job weights. When the leader’s weights are arbitrary, the problem is NP-hard. However, when all the jobs are of unit weight for the leader, we provide a heuristic algorithm based on iterative LP-rounding along with computational results, and provide a sufficient condition when the LP-solution is integral. In addition, if the follower weights induce a monotone (increasing or decreasing) processing time order in any optimal solution, the problem becomes polynomially solvable. As a by-product, we characterise a new polynomially solvable special case of the MAX m-CUT problem, and provide a new linear programming formulation for the P||?_j C_j{P||\sum_j C_j} problem. Finally, we present some results on the bilevel order acceptance problem, where the leader decides on the acceptance of orders and the follower sequences the jobs. Each job has a deadline and if a job is accepted, it cannot be late. The leader’s objective is to maximise the total weight of accepted jobs, whereas the follower aims at minimising the total weighted job completion times. For this problem, we generalise some known single-level machine scheduling algorithms.     

Solving the n-job 3-stage flexible flowshop scheduling problem using an agent-based approach

In this paper, a general methodology of agent-based manufacturing systems scheduling, incorporating game theoretic analysis of agent cooperation is presented to solve the n-job 3-stage flexible flowshop scheduling problem. The flowshops are flexible in the sense that a job can be processed by any of the identical machines at each stage. Our objective is to schedule a set of n jobs so as to minimize the makespan. We perform error bound analysis using the lower bound estimates developed in the literature as a datum for comparing the agent-based scheduling solutions with other heuristic solutions. The results of the evaluation show that the agent-based scheduling approach outperforms existing heuristics for the majority of the testing problems.     

A Branch-and-Bound Algorithm for the Continuous-Process Job-Shop Scheduling Problem

The continuous-process job-shop scheduling problem (CPJS) arises typically in the following way: (1) a set of M machines or production facilities are available; (2) a set of N jobs are to be processed through these machines in accordance with a technological matrix; (3) the machines associated with a given job must all be used simultaneously for the completion of this job; (4) a predetermined production time is required for each job; (5) the objective is to determine a production schedule which minimizes the total completion time (makespan) of all jobs. A branch-and-bound type algorithm for the solution of the (CPJS) problem is presented.     

JOB ASSIGNMENT ALGORITHMS ON UNIFORM PARALLEL MACHINES

Uniform parallel machine scheduling problems with a makespan measure cannot generally be solved within polynomial time complexity. This paper considers special problems with a single type of job on the uniform parallel machines, where each machine is available at a given ready time. Also the machine can be restricted on the number of jobs to be processed. The objective is to develop job assignment or batching algorithms which minimize makespan. When all the machines are available at time zero and have no restriction on the number of assignable jobs, a lower bound and optimal solution properties are derived. Based upon these properties, a polynomial algorithm is suggested to find the optimal job assignment on each machine. Three generalized problems are considered under the following situations: (1) some machines have capacity restrictions on the production batch, (2) each machine has its ready time, and (3) the jobs require series-parallel operations. The generalized problems arc also characterized and polynomial algorithms are developed for the same aim of optimal job assignment, except for the case of series-parallel operations. A heuristic algorithm is suggested with numerical tests for the series-parallel operations problem     

Two-stage hybrid flow shop scheduling with preventive maintenance using multi-objective tabu search method

This paper investigates an integrated bi-objective optimisation problem with non-resumable jobs for production scheduling and preventive maintenance in a two-stage hybrid flow shop with one machine on the first stage and m identical parallel machines on the second stage. Sequence-dependent set-up times and preventive maintenance (PM) on the first stage machine are considered. The scheduling objectives are to minimise the unavailability of the first stage machine and to minimise the makespan simultaneously. To solve this integrated problem, three decisions have to be made: determine the processing sequence of jobs on the first stage machine, determine whether or not to perform PM activity just after each job, and specify the processing machine of each job on the second stage. Due to the complexity of the problem, a multi-objective tabu search (MOTS) method is adapted with the implementation details. The method generates non-dominated solutions with several parallel tabu lists and Pareto dominance concept. The performance of the method is compared with that of a well-known multi-objective genetic algorithm, in terms of standard multi-objective metrics. Computational results show that the proposed MOTS yields a better approximation.     

Scheduling a tempered glass manufacturing system: a three-stage hybrid flow shop model

Chinese tempered glass has entered a fast and stable growing era. To improve the productivity of tempered glass manufacturers, this paper investigates a scheduling problem in tempered glass production system, originated from a tempered glass manufacturer in China. This problem can be formulated as a three-stage hybrid flow shop (HFS). Single and batch processing machines coexist in this HFS. Besides, a limited buffer, between the first two stages, and machine eligibility requirement are also significant characteristics. To address this complicated scheduling problem, we first establish an integer programming model with the objective of minimising the makespan, i.e. the maximum completion time of jobs in the system. Due to the strong NP-hard nature of the problem, we then propose a constructive heuristic method, a genetic algorithm, as well as a simulated annealing algorithm, to solve practical large-scale problems. Computational results demonstrate the efficiency of the proposed approaches.     

Quantum-inspired ant colony optimisation algorithm for a two-stage permutation flow shop with batch processing machines

This paper studied two-stage permutation flow shop problems with batch processing machines, considering different job sizes and arbitrary arrival times, with the optimisation objective of minimising the makespan. The quantum-inspired ant colony optimisation (QIACO) algorithm was proposed to solve the problem. In the QIACO algorithm, the ants are divided into two groups: one group selects the largest job in terms of job size as the initial job for each batch and the other group selects the smallest job as the initial job for each batch. Each group of ants has its own pheromone matrix. In the computational experiment, our novel algorithm was compared with the hybrid discrete differential evolution (HDDE) algorithm and the batch-based hybrid ant colony optimisation (BHACO) algorithm. Although the HDDE algorithm has a shorter run time, the quality of the solution for large-scale jobs is not good, while the BHACO algorithm always obtains a better solution but requires a longer run time. The computational results show that the QIACO algorithm embedded in the quantum information has advantages in terms of both solution quality and running time.     

Scheduling two identical parallel machines with preparation constraints

In this paper, we consider the problem of scheduling a set of n jobs on two identical machines with preparation constraints. Each job requires before its execution a set of resources and a non-negligible preparation time. The objective is to minimise the makespan. This problem is NP-hard. We prove the NP-hardness of two specific cases where in the first case preparation times take only three values, whereas in the second case preparation times and the release dates take only two values, respectively. Then, we present some special cases and heuristic algorithms along with an experimental study.     

Scheduling with multi-attribute set-up times on unrelated parallel machines

This paper studies a problem in the knitting process of the textile industry. In such a production system, each job has a number of attributes and each attribute has one or more levels. Because there is at least one different attribute level between two adjacent jobs, it is necessary to make a set-up adjustment whenever there is a switch to a different job. The problem can be formulated as a scheduling problem with multi-attribute set-up times on unrelated parallel machines. The objective of the problem is to assign jobs to different machines to minimise the makespan. A constructive heuristic is developed to obtain a qualified solution. To improve the solution further, a meta-heuristic that uses a genetic algorithm with a new crossover operator and three local searches are proposed. The computational experiments show that the proposed constructive heuristic outperforms two existed heuristics and the current scheduling method used by the case textile plant.     

含不一致任务重量的同型熔炼炉批调度优化

为求解含不一致任务重量的同型熔炼炉批调度问题,建立了最小化最大任务完工时间优化模型,设计了一种混合粒子群算法（HPSO）。算法使用随机生成的任务序列作为粒子,采用批首次匹配（BFF）规则对任务序列分批,最长加工时间（LPT）规则将批分配到批处理机,并提出了一种最小完工时间差（MCD）规则对LPT调度结果进行优化;为避免早熟,算法引入交叉和变异操作搜索最优解。通过仿真实验与SA、GA算法对比,实验结果表明算法具有良好的性能。     

Meta-heuristics for the distributed two-stage assembly scheduling problem with bi-criteria of makespan and mean completion time

This paper focuses on the distributed two-stage assembly flowshop scheduling problem for minimising a weighted sum of makespan and mean completion time. This problem involves two inter-dependent decision sub-problems: (1) how to allocate jobs among factories and (2) how to schedule the assigned jobs at each factory. A mathematical model is formulated for solving the small-sized instances of the problem. Since the NP-hardness of the problem, we also proposed a variable neighbourhood search (VNS) algorithm and a hybrid genetic algorithm combined with reduced variable neighbourhood search (GA-RVNS) to solve the distributed two-stage assembly flowshop scheduling problems and approximately optimise makespan and mean completion time simultaneously. Computational experiments have been conducted to compare the performances of the model and proposed algorithms. For a set of small-sized instances, both the model and the proposed algorithms are effective. The proposed algorithms are further evaluated on a set of large-sized instances. The results statistically show that both GA-RVNS and VNS obtain much better performances than the GA without RVNS-based local search step (GA-NOV). For the instances with small numbers of jobs, VNS achieves better performances than GA-RVNS. However, for the instances with large numbers of jobs, GA-RVNS yields better performances than the VNS. It is also shown that the overall performances of VNS are very close to GA-RVNS with different numbers of factories, weights given to makespan and numbers of machines at the first stage.     

Two-machine flow shop problem with unit-time operations and conflict graph

This paper addresses the problem of scheduling, on a two-machine flow shop, a set of unit-time operations subject to the constraints that some conflicting jobs cannot be scheduled simultaneously on different machines. In the context of our study, these conflicts are modelled by general graphs. The problem of minimising the maximum completion time (makespan) is known to be NP-hard in the strong sense. We propose a mixed-integer linear programming (MILP) model. Then, we develop a branch and bound algorithm based on new lower and upper bound procedures. We further provide a computer simulation to measure the performance of the proposed approaches. The computational results show that the branch and bound algorithm outperforms the MILP model and can solve instances of size up to 20,000 jobs.     

A multi-point simulated annealing heuristic for solving multiple objective unrelated parallel machine scheduling problems

This study considers the problem of job scheduling on unrelated parallel machines. A multi-objective multi-point simulated annealing (MOMSA) algorithm was proposed for solving this problem by simultaneously minimising makespan, total weighted completion time and total weighted tardiness. To assess the performance of the proposed heuristic and compare it with that of several benchmark heuristics, the obtained sets of non-dominated solutions were assessed using four multi-objective performance indicators. The computational results demonstrated that the proposed heuristic markedly outperformed the benchmark heuristics in terms of the four performance indicators. The proposed MOMSA algorithm can provide a new benchmark for future research related to the unrelated parallel machine scheduling problem addressed in this study.     

Hybrid flowshop scheduling with family setup time and inconsistent family formation

This research considers a hybrid flowshop scheduling problem where jobs are organised in families according to their machine settings and tools. The family setup time arises when a machine shifts from processing one job family to another. The problem is compounded by the challenges that the formation of job families is different in different stages and only a limited number of jobs can be processed within one setup. This type of problem is common in the production process of standard metal components. This paper aims to propose two approaches to solve this problem. One is a metaheuristic in the form of a genetic algorithm and the other is a heuristic. The proposed approaches are compared and contrasted against the two relevant metaheuristic and heuristic adapted from solving a generalised sequence-dependent setup flowshop problem. Comparative results indicate that the proposed genetic algorithm has better performance on minimising makespan and the heuristic is more effective on reducing family setup time.     

Genetic algorithm with local search for the unrelated parallel machine scheduling problem with sequence-dependent set-up times

In this paper, a genetic algorithm (GA) with local search is proposed for the unrelated parallel machine scheduling problem with the objective of minimising the maximum completion time (makespan). We propose a simple chromosome structure consisting of random key numbers in a hybrid genetic-local search algorithm. Random key numbers are frequently used in GAs but create additional difficulties when hybrid factors are implemented in a local search. The best chromosome of each generation is improved using a local search during the algorithm, but the better job sequence (which might appear during the local search operation) must be adapted to the chromosome that will be used in each successive generation. Determining the genes (and the data in the genes) that would be exchanged is the challenge of using random numbers. We have developed an algorithm that satisfies the adaptation of local search results into the GAs with a minimum relocation operation of the genes’ random key numbers – this is the main contribution of the paper. A new hybrid approach is tested on a set of problems taken from the literature, and the computational results validate the effectiveness of the proposed algorithm.