A memetic differential evolution algorithm for energy-efficient parallel machine scheduling

This paper considers an energy-efficient bi-objective unrelated parallel machine scheduling problem to minimize both makespan and total energy consumption. The parallel machines are speed-scaling. To solve the problem, we propose a memetic differential evolution (MDE) algorithm. Since the problem involves assigning jobs to machines and selecting an appropriate processing speed level for each job, we characterize each individual by two vectors: a job-machine assignment vector and a speed vector. To accelerate the convergence of the algorithm, only the speed vector of each individual evolves and a list scheduling heuristic is applied to derive its job-machine assignment vector based on its speed vector. To further enhance the algorithm, we propose efficient speed adjusting and job-machine swap heuristics and integrate them into the algorithm as a local search approach by an adaptive meta-Lamarckian learning strategy. Computational results reveal that the incorporation of list scheduling heuristic and local search greatly strengthens the algorithm. Computational experiments also show that the proposed MDE algorithm outperforms SPEA-II and NSGA-II significantly.     

基于改进遗传算法的置换装配线调度问题研究

企业的置换装配线调度问题（Permutation Assembly-line Scheduling Problem,PASP）是一类典型的NP-hard型生产调度问题,是现代集成制造系统CIMS极为关心的问题。该问题可以具体描述为n个工件要在m台机器上加工,每个工件需要经过m道工序,每道工序要求不同的机器,这n个工件通过m台机器的顺序相同,它们在每台机器上的加工顺序也相同,问题的主要目标是找到n个工件在每台机器上的最优加工顺序,使得最大完工时间最小。由于PASP问题的NP-hard性质,本文使用遗传算法对其进行求解。尽管遗传算法常用以求解调度问题,但其选择与交叉机制易导致局部最优及收敛慢。因此,本文提出基于区块挖掘与重组的改进遗传算法用于求解置换装配线调度问题。首先通过关联规则挖掘出不同的优秀基因,然后将具有较优结果的基因组合为优势区块,产生具优势的人工解,并引入高收敛性的局部搜索方法,提高搜索到最优解的机会与收敛效率。本文以OR-Library中Taillard标准测试例来验证改进遗传算法的求解质量与效率,结果证明：本文所提算法与其它求解调度问题的现有5种知名算法相比,不仅收敛速度较快,同时求解质量优于它们。     

O(m2) ALGORITHMS FOR THE TWO AND THREE SUBLOT LOT STREAMING PROBLEM

Lot streaming is the process of splitting a job or lot into sublots to reduce its makespan on a sequence of machines. The goal in the lot streaming problem is to find the optimal size of each sublot that will minimize the makespan. The makespan is defined as the time the last sublot completes its processing on the last machine. If the sizes of these sublots are restricted to remain the same on all machines, the solution is called a consistent sublot solution. However, if the sizes of the sublots are allowed to vary, the solution is referred to as a nonconsistent or variable sublot solution. Also, if the machines must be in operation continuously from the first to the last sublot, the solution is a no idling solution. When setups are explicitly considered in the problem, there will be two cases. If setups on each machine require some portion of the first sublot be present by the machine, the problem is referred to as the attached setup time problem. If setups can be performed ahead of time before the first sublot reaches the particular machine, the corresponding problem is referred to as the detached setup problem. Finally, if the machines are allowed to be idle between the processing of sublots, the resultant solution is an intermittent idling solution. In this paper, the consistent sublot lot streaming problem with intermittent idling and no setups is discussed. The models developed also assume that the number of sublots are fixed and known. The m machine two sublot lot streaming problem is reviewed. An algorithm for the three sublot, m machine problem is derived using a network representation of the problem. The complexity of the algorithm is O (m²). Finally, using the insights from three sublot problem, a heuristic algorithm is provided for the m machine, n sublot problems. The results on the proposed heuristic are very encouraging; average percent deviation from optimal makespan is approximately at 0.76% on 155 randomly generated problems with different m and n values.     

The flexible blocking job shop with transfer and set-up times

The Flexible Blocking Job Shop (FBJS) considered here is a job shop scheduling problem characterized by the availability of alternative machines for each operation and the absence of buffers. The latter implies that a job, after completing an operation, has to remain on the machine until its next operation starts. Additional features are sequence-dependent transfer and set-up times, the first for passing a job from a machine to the next, the second for change-over on a machine from an operation to the next. The objective is to assign machines and schedule the operations in order to minimize the makespan. We give a problem formulation in a disjunctive graph and develop a heuristic local search approach. A feasible neighborhood is constructed, where typically a critical operation is moved (keeping or changing its machine) together with some other operations whose moves are “implied”. For this purpose, we develop the theoretical framework of job insertion with local flexibility, based on earlier work of Gröflin and Klinkert on insertion. A tabu search that consistently generates feasible neighbor solutions is then proposed and tested on a larger test set. Numerical results support the validity of our approach and establish first benchmarks for the FBJS.     

考虑分时电价的多目标批调度问题蚁群算法求解

对同时优化电力成本和制造跨度的多目标批处理机调度问题进行了研究,设计了两种多目标蚁群算法,基于工件序的多目标蚁群算法(J-PACO,Job-based Pareto Ant Colony Optimization)和基于成批的多目标蚁群算法(B-PACO,Batch-based Pareto Ant Colony Optimization)对问题进行求解分析。由于分时电价中电价是时间的函数,因而在传统批调度进行批排序的基础上,需要进一步确定批加工时间点以测定电力成本。提出的两种蚁群算法分别将工件和批与时间线相结合进行调度对此类问题进行求解。通过仿真实验将两种算法对问题的求解进行了比较,仿真实验表明B-PACO算法通过结合FFLPT(First Fit Longest Processing Time)启发式算法先将工件成批再生成最终方案,提高了算法搜索效率,并且在衡量算法搜索非支配解数量的Q指标和衡量非支配集与Pareto边界接近程度的HV指标上,均优于J-PACO算法。     

On the optimality of the TLS algorithm for solving the online-list scheduling problem with two job types on a set of multipurpose machines

In this paper we study the optimality of the TLS algorithm for solving the online scheduling problem of minimizing the makespan on a set of m multipurpose machines, where there are two different job types and each job type can only be processed on a unique subset of machines. The literature shows that the TLS algorithm is optimal for the special cases where either m=2 or where all processing times are restricted to unity. We show that the TLS algorithm is optimal also for the special cases where the job processing times are either job type or machine set dependent. For both cases, the optimality of the TLS algorithm is proven by showing that its competitive ratio matches the lower bound for any processing set and processing time parameters.     

The blocking job shop with rail-bound transportation

The blocking job shop with rail-bound transportation (BJS-RT) considered here is a version of the job shop scheduling problem characterized by the absence of buffers and the use of a rail-bound transportation system. The jobs are processed on machines and are transported from one machine to the next by mobile devices (called robots) that move on a single rail. The robots cannot pass each other, must maintain a minimum distance from each other, but can also “move out of the way”. The objective of the BJS-RT is to determine for each machining operation its starting time and for each transport operation its assigned robot and starting time, as well as the trajectory of each robot, in order to minimize the makespan. Building on previous work of the authors on the flexible blocking job shop and an analysis of the feasible trajectory problem, a formulation of the BJS-RT in a disjunctive graph is derived. Based on the framework of job insertion in this graph, a local search heuristic generating consistently feasible neighbor solutions is proposed. Computational results are presented, supporting the value of the approach.     

Minimization of maximum lateness under linear deterioration

This paper considers a single-machine scheduling problem to minimize the maximum lateness. The processing time of each job is a linear function of the time when the job starts processing. This problem is known to be -hard in the literature. In this paper, we design a branch-and-bound algorithm for deriving exact solutions by incorporating several properties concerning dominance relations and lower bounds. These properties produce synergic effects in accelerating the solution finding process such that the algorithm can solve problems of 100 jobs within 1 min on average. To compose approximate solutions, we revise a heuristic algorithm available in the literature and propose several hybrid variants. Numerical results evince that the proposed approaches are very effective in successfully reporting optimal solutions for most of the test cases.     

一种差异工件单机批调度问题的蚁群优化算法

由于在利用蚁群算法构建差异工件(即工件有尺寸差异)单机批调度问题的解时,批的加工时间是不确定的.从而不能类似于经典调度问题的蚁群算法把批加工时间的倒数作为蚁群算法中的启发式信息,引入批的利用率和批的负载均衡率作为蚁群算法中的启发式信息,提出了JACO(ant colony optimization based a job sequence)和BACO(ant colony optimization based a batch sequence)两种蚁群优化算法.在算法JACO中,解的编码为工件序列,它对应着用BF(best fit)分批规则生成的调度方案,信息素代表工件间的排列顺序;在算法BACO中,解的编码为批序列,信息素代表工件间的批相关性,由此信息素通过中间信息素量来构造相应的解,并引入特定的局部优化策略,提高了算法的搜索效率.实验表明,与以往文献中的SA(simula-ted annealing)、GA(genetic algorithm)算法以及FFLPT(first-fit longest processing time)、BFLPT (best-fit longest processing time)启发式规则相比,算法JACO和BACO明显优于它们,且BACO算法比JACO算法效果更好.     

A hybrid genetic algorithm/mathematical programming approach to the multi-family flowshop scheduling problem with lot streaming

This paper presents a hybrid genetic algorithm/mathematical programming heuristic for the n-job, m-machine flowshop problems with lot streaming. The number of sublots for each job and the size of sublots are directly addressed by the heuristic and setups may be sequence-dependent. A new aspect of the problem, the interleaving of sublots from different jobs in the processing sequence, is developed and addressed. Computational results from 12 randomly generated test sets of 24 problems each are presented.     

Improved approximation algorithms for non-preemptive multiprocessor scheduling with testing

Multiprocessor scheduling, also called scheduling on parallel identical machines to minimize the makespan, is a classic optimization problem which has been extensively studied. Scheduling with testing is an online variant, where the processing time of a job is revealed by an extra test operation, otherwise the job has to be executed for a given upper bound on the processing time. Albers and Eckl recently studied the multiprocessor scheduling with testing; among others, for the non-preemptive setting they presented an approximation algorithm with competitive ratio approaching 3.1016 when the number of machines tends to infinity and an improved approximation algorithm with competitive ratio approaching 3 when all test operations take one unit of time each. We propose to first sort the jobs into non-increasing order of the minimum value between the upper bound and the testing time, then partition the jobs into three groups and process them group by group according to the sorted job order. We show that our algorithm achieves better competitive ratios, which approach 2.9513 when the number of machines tends to infinity in the general case; when all test operations each takes one time unit, our algorithm achieves even better competitive ratios approaching 2.8081.

     

An exchange heuristic for resource constrained scheduling with consideration given to opportunities for parallel processing

This paper describes a heuristic which produces efficient makespans for resource-constrained scheduling problems with parallel processing capabilities. This heuristic was initially developed for the scheduling of army battalion training exercises. The original heuristic has also been successfully applied to solve problems in project scheduling with limited resources, generalized job shop scheduling, and resource-constrained scheduling. The exchange heuristic requires an initial feasible solution upon which it improves the makespan by efficiently and systematically shuffling activities while maintaining feasibility. The method has recently been modified twice, termed the intelligent version and naive version, respectively, such that its ability to reduce the initial makespan is enhanced. In this study     

A hybrid approach to sequencing jobs using heuristic rules and neural networks

A hybrid approach to solve job sequencing problems using heuristic rules and artificial neural networks is proposed. The problem is to find a job sequence for a single machine that minimizes the total weighted tardiness of the jobs. Two different cases are considered: (1) when there are no setups, and (2) when there are sequence-dependent setup times. So far, successful heuristic rules for these cases are: apparent tardiness cost (ATC) rule proposed by Vepsalainen and Morton for the former case, and an extended version of the ATC rule (ATCS) proposed by Lee, Bhaskaran, and Pinedo for the latter. Both approaches utilize some look-ahead parameters for calculating the priority index of each job. As reported by Bhaskaran and Pinedo, the proper value of the look-ahead parameter depends upon certain problem characteristics, such as due-date tightness and due-date range. Thus, an obvious extension of the ATC or the ATCS rule is to adjust the parameter values depending upon the problem characteristics: this is known to be a difficult task. In this paper, we propose an application of a neural network as a tool to ‘predict’ proper values of the look-ahead parameters. Our computational tests show that the proposed hybrid approach outperforms both the ATC rule with a fixed parameter value and the ATCS using the heuristic curve-fitting method.     

An effective iterated greedy algorithm for the mixed no-idle permutation flowshop scheduling problem

In the no-idle flowshop, machines cannot be idle after finishing one job and before starting the next one. Therefore, start times of jobs must be delayed to guarantee this constraint. In practice machines show this behavior as it might be technically unfeasible or uneconomical to stop a machine in between jobs. This has important ramifications in the modern industry including fiber glass processing, foundries, production of integrated circuits and the steel making industry, among others. However, to assume that all machines in the shop have this no-idle constraint is not realistic. To the best of our knowledge, this is the first paper to study the mixed no-idle extension where only some machines have the no-idle constraint. We present a mixed integer programming model for this new problem and the equations to calculate the makespan. We also propose a set of formulas to accelerate the calculation of insertions that is used both in heuristics as well as in the local search procedures. An effective iterated greedy (IG) algorithm is proposed. We use an NEH-based heuristic to construct a high quality initial solution. A local search using the proposed accelerations is employed to emphasize intensification and exploration in the IG. A new destruction and construction procedure is also shown. To evaluate the proposed algorithm, we present several adaptations of other well-known and recent metaheuristics for the problem and conduct a comprehensive set of computational and statistical experiments with a total of 1750 instances. The results show that the proposed IG algorithm outperforms existing methods in the no-idle and in the mixed no-idle scenarios by a significant margin.     

Processing in a multi-machine batch manufacturing system with sequence dependent cost

Many workcells in batch manufacturing systems are populated with multiple, nonidentical machines that perform similar tasks. Because of the size of a batch when a job arrives, it may be uneconomical to set up two or more machines to process the same job simultaneously. An economic decision has to be made as regards which machine in the cell to assign the job. Likewise, many multi-operation jobs can be processed using one of several feasible operation sequences that may lead to different total manufacturing costs. The cost differences are the result of several factors, among which are processing time and cost dependencies between operations, fixturing requirements, and material handling requirements. When the workcell machine selection decision is considered along with the operation sequencing decision, determination of the best machine in a cell and the best operation sequence for the batch is a non-trivial task. In this paper, we address the problem of selecting the best machine within a cell and the best operation sequence for a batch when operation cost is machine and sequence dependent. The problem is modeled mathematically and solved using a heuristic algorithm. The performance of the algorithm is compared with that of an exact solution procedure.     

A heuristic algorithm for minimizing maintenance workforce level

We develop a heuristic algorithm for minimizing the workforce level required to accommodate all the maintenance jobs requested within a specific time interval. Each maintenance job has its own release and due dates as well as the required man-days, and must be scheduled in a noninterrupted time interval, i.e. without preemption. However, the duration of each job is not fixed, but to be determined within a specific range. We show that this problem can be seen as a variant of the two-dimensional bin-packing problem with some additional constraints. We develop a non-linear mixed integer programming model for the proposed problem, and employ some well-known bin-packing algorithms to develop an efficient heuristic algorithm. In order to evaluate the performance of the proposed heuristic, we present a computationally efficient scheme for getting a good lower bound for the actual minimum. The computational experiment shows that the proposed heuristic algorithm performs quite satisfactorily in practice.     

Job sequencing and component set-up on a surface mount placement machine

In this paper, the problem of sequencing PCB assembly jobs on an automatic SMD placement machine is addressed. The objective is to minimize the makespan. Moreover, both the job arrival times and the precedence constraints for those jobs requiring component placement on the primary and secondary side of the same board must be taken into account. A considerable set-up time occurs when switching from one board type to another, which depends on the number of component feeders to be replaced in the magazine of the assembly machine. The exchange of component feeders is complicated by the fact that each feeder occupies a different number of magazine positions. Theoretically, the minimum makespan required for a given batch of jobs could be derived by solving the order sequencing and the component set-up problems simultaneously. However, optimal solutions are practically unattainable for problems of realistic size. Therefore, efficient heuristic solution procedures are developed which exploit component commonality between PCB types. The numerical results obtained indicate the practicality of the proposed heuristics in an industrial application.     

Performance analysis of re-entrant flow shop with single-job and batch machines using mean value analysis

We propose an approximate method based on the mean value analysis for estimating the average performance of re-entrant flow shop with single-job machines and batch machines. The main focus is on the steady-state averages of the cycle time and the throughput of the system. Characteristics of the re-entrant flow and inclusion of the batch machines complicate the exact analysis of the system. Thus, we propose an approximate analytic method for obtaining the mean waiting time at each buffer of the workstation and a heuristic method to improve the result of the analytic method. We compare the results of the proposed approach with a simulation study using some numerical examples.     

Semi-online scheduling on 2 machines under a grade of service provision with bounded processing times

We study the problem of semi-online scheduling on 2 machines under a grade of service (GoS). GoS means that some jobs have to be processed by some machines to be guaranteed a high quality. The problem is online in the sense that jobs are presented one by one, and each job shall be assigned to a time slot on its arrival. Assume that the processing time p _i of every job J _i is bounded by an interval [a,α a], where a>0 and α>1 are two constant numbers. By knowing the bound of jobs’ processing times, we denote it by semi-online problem. We deal with two semi-online problems.     

A multiobjective genetic algorithm for job shop scheduling

In this paper, a Multi Objective Genetic Algorithm (MOGA) is proposed to derive the optimal machine-wise priority dispatching rules ( pdrs ) to resolve the conflict among the contending jobs in the Giffler and Thompson (GT) procedure applied for job shop problems. The performance criterion considered is the weighed sum of the multiple objectives minimization of makespan, minimization of total idle time of machines and minimization of total tardiness. The weights assigned for combining the objectives into a scalar fitness function are not constant. They are specified randomly for each evaluation. This in turn leads to the multidirectional search in the proposed MOGA, which in turn mitigates the solution being entrapped in local minima. The applicability and usefulness of the proposed methodology for the scheduling of job shops is illustrated with 28 benchmark problems available in the open literature.