Multi-objective genetic algorithm and its applications to flowshop scheduling

In this paper, we propose a multi-objective genetic algorithm and apply it to flowshop scheduling. The characteristic features of our algorithm are its selection procedure and elite preserve strategy. The selection procedure in our multi-objective genetic algorithm selects individuals for a crossover operation based on a weighted sum of multiple objective functions with variable weights. The elite preserve strategy in our algorithm uses multiple elite solutions instead of a single elite solution. That is, a certain number of individuals are selected from a tentative set of Pareto optimal solutions and inherited to the next generation as elite individuals. In order to show that our approach can handle multi-objective optimization problems with concave Pareto fronts, we apply the proposed genetic algorithm to a two-objective function optimization problem with a concave Pareto front. Last, the performance of our multi-objective genetic algorithm is examined by applying it to the flowshop scheduling problem with two objectives: to minimize the makespan and to minimize the total tardiness. We also apply our algorithm to the flowshop scheduling problem with three objectives: to minimize the makespan, to minimize the total tardiness, and to minimize the total flowtime.     

A high performing metaheuristic for multi-objective flowshop scheduling problem

Genetic algorithm is a powerful procedure for finding an optimal or near optimal solution for the flowshop scheduling problem. This is a simple and efficient algorithm which is used for both single and multi-objective problems. It can easily be utilized for real life applications. The proposed algorithm makes use of the principle of Pareto solutions. It mines the Pareto archive to extract the most repetitive sequences, and constitutes artificial chromosome for generation of the next population. In order to guide the search direction, this approach coupled with variable neighborhood search. This algorithm is applied on the flowshop scheduling problem for minimizing makespan and total weighted tardiness. For the assessment of the algorithm, its performance is compared with the MOGLS [1]. The results of the experiments allow us to claim that the proposed algorithm has a considerable performance in this problem.     

An adaptive multi-population genetic algorithm to solve the multi-objective group scheduling problem in hybrid flexible flowshop with sequence-dependent setup times

In this paper we consider a multi-objective group scheduling problem in hybrid flexible flowshop with sequence-dependent setup times by minimizing total weighted tardiness and maximum completion time simultaneously. Whereas these kinds of problems are NP-hard, thus we proposed a multi-population genetic algorithm (MPGA) to search Pareto optimal solution for it. This algorithm comprises two stages. First stage applies combined objective of mentioned objectives and second stage uses previous stage’s results as an initial solution. In the second stage sub-population will be generated by re-arrangement of solutions of first stage. To evaluate performance of the proposed MPGA, it is compared with two distinguished benchmarks, multi-objective genetic algorithm (MOGA) and non-dominated sorting genetic algorithm II (NSGA-II), in three sizes of test problems: small, medium and large. The computational results show that this algorithm performs better than them.     

一种基于模拟退火的多目标Memetic算法

为了改善多目标进化算法的搜索效率，提出了基于模拟退火的多目标Memetic算法．此算法根据Pareto占优关系评价个体适应值，采用模拟退火进行局部搜索，并结合交叉算子和基于网格密度的选择机制改善算法的收敛速度和解的均衡分布．flowshop调度问题算例的仿真结果表明，基于模拟退火的多目标Memetic算法能够产生更接近Pareto前沿的近似集．     

Multi-objective colonial competitive algorithm for hybrid flowshop problem

This study analyses the multi-objective optimization in hybrid flowshop problem, in which two conflicting objectives, makespan and total weighted tardiness, are considered to be minimized simultaneously. The multi-objective version of Colonial Competitive Algorithm (CCA) for real world optimization problem is introduced and investigated. In contrast to multi-objective problems solved by CCA, presented in the literature, which used the combination of the objectives as single objective, the proposed algorithm is established on Pareto solutions concepts. Another novelty of this paper is estimating the power of each imperialist by a probabilistic criterion for this multi objective algorithm. Besides that, the variable neighborhood search is implemented as an assimilation strategy. Performance of the algorithm is finally compared with a famous algorithm for scheduling problem, NSGA-II, and the multi-objective form of CCA [28].     

Supply chain-oriented two-stage assembly flowshops with sequence-dependent setup times

Supply chain-oriented scheduling problems have received recent recognition among production research scholars due to their ability in integrating production planning and control across manufacturing systems. This study contributes to the literature of the distributed scheduling problems developing an original Mixed-Integer Linear Programming (MILP) formulation to the Distributed Two-Stage Assembly Flowshop Scheduling Problem with Sequence-Dependent Setup Times (DTSAFSP-SDSTs). Besides, the Iterated Greedy algorithm is extended to effectively solve this relatively complex production scheduling situation considering the makespan as the optimization criterion. Extensive numerical tests and statistical analyses are conducted to evaluate the effectiveness of the developed solution algorithm. Results showed that the Improved Iterated Greedy (IIG) algorithm yields the best solution in nearly all of the large-scale instances. The statistical test of significance confirmed that IIG is superior to the current-best-performing algorithm. This study contributes to the transition from standalone optimization to integrated production planning and control of distributed manufacturing systems.     

云环境中基于分解的多目标工作流调度算法

云服务提供商在给用户提供海量虚拟资源的同时,也面临着一个现实的问题,即怎样调度这些资源,以最小的代价(完工时间、执行费用、资源利用率等)完成工作流的执行。针对IaaS环境下的工作流调度问题,以完工时间和执行费用作为目标,提出了一种基于分解的多目标工作流调度算法。该算法结合了基于列表的启发式算法和多目标进化算法的选择过程,采用一种分解方法,将多目标优化问题分解为一组单目标优化子问题,然后同时求解这些单目标子问题,使得调度过程更为简单有效。算法利用天马项目发布的现实世界中的工作流进行实验,结果表明,和MOHEFT算法以及NSGA-II*算法相比较,所提出的算法能得到更优的Pareto解集,同时具有更低的时间复杂度。     

改进人工蜂群求解多目标柔性作业车间调度问题

针对多个目标约束的柔性作业车间问题,本文采用基于Pareto解集的改进离散人工蜂群算法来求解.由于经典人工蜂群算法的选择概率不适用于多目标问题,本文对选择概率进行了重定义,将排序引入选择概率中;同时采用基于变异操作的邻域搜索方法进行局部搜索,并使用混合列交叉算子提高种群的多样性;采用Harmonic平均距离对Pareto解集进行裁剪,完成对Pareto解集的更新.最后通过实例测试及仿真实验,验证了本文算法在求解多目标柔性作业车间调度时的有效性.     

不同支配关系的多目标算法的柔性作业调度

为了提高多目标进化算法所获得解的质量,研究者做了大量的研究,传统的基于Pareto支配关系的多目标进化算法具有一定的局限性。本文利用不同的支配关系与NSGA-II(Non-dominated Sorting Genetic Algorithm)算法相结合,对单机器人搬运的柔性作业车间调度的多目标优化问题进行求解,通过实验比较分析了不同方法在多目标优化问题求解中的优劣性。本文以NSGA-II为框架结合Lorenz支配关系和CDAS(Control Dominance Area of Solutions)支配关系并与传统的基于Pareto支配关系的NSGA-II三种算法去研究同一优化调度问题,发现基于Lorenz支配关系和CDAS支配关系的优化算法比基于传统的Pareto支配关系的优化算法的效果更佳。     

Bi-objective scheduling for reentrant hybrid flow shop using Pareto genetic algorithm

This paper deals with a scheduling problem for reentrant hybrid flowshop with serial stages where each stage consists of identical parallel machines. In a reentrant flowshop, a job may revisit any stage several times. Local-search based Pareto genetic algorithms with Minkowski distance-based crossover operator is proposed to approximate the Pareto optimal solutions for the minimization of makespan and total tardiness in a reentrant hybrid flowshop. The Pareto genetic algorithms are compared with existing multi-objective genetic algorithm, NSGA-II in terms of the convergence to optimal solution, the diversity of solution and the dominance of solution. Experimental results show that the proposed crossover operator and local search are effective and the proposed algorithm outperforms NSGA-II by statistical analysis.     

A genetic algorithm with genes-association recognition for flowshop scheduling problems

In this paper, we consider a flowshop scheduling problem with a special blocking RCb (Release when Completing Blocking). This flexible production system is prevalent in some industrial environments. Genetic algorithms are first proposed for solving these flowshop problems and different initial populations have been tested to find which is best adapted. Then, a method is proposed for further improving genetic algorithm found solutions, which consists in marking out recurrent genes association occurrences in an already genetic algorithm optimized population. This idea directly follows Holland’s first statement about nature observations. Here, proposed idea is that populations well adapted to a problem have an adapted genetic code with common properties. We propose to mark out these properties in available genetic code to further improve genetic algorithm efficiency. Implementation of this method is presented and obtained results on flowshop scheduling problems are discussed.     

Optimizing distributed no-wait flow shop scheduling problem with setup times and maintenance operations via iterated greedy algorithm

Nowadays, distributed scheduling problem is a reality in many companies. Over the last years, an increasingly attention has been given to the distributed flow shop scheduling problem and the addition of constraints to the problem. This article introduces the distributed no-wait flow shop scheduling problem with sequence-dependent setup times and maintenance operations to minimize makespan. A mixed-integer linear programming (MILP) is to mathematically describe the problem and heuristic procedures to incorporate maintenance operations to job scheduling are proposed. An Iterated Greedy with Variable Search Neighborhood (VNS), named IG_NM, is proposed to solve small and large instances with size of 4,800 and 13,200 problems, respectively. Computational experiments were conducted to evaluate the performance of IG_NM in comparison with MILP and the most recent methods of literature of distributed flow shop scheduling problems. Statistical results show that in the trade-off between effectiveness and efficiency the proposed IG_NM outperformed other metaheuristics of the literature.     

求解多目标作业车间调度问题的混合变异杂草优化算法

针对多目标作业车间调度问题,提出一种混合变异杂草优化算法。该算法采用基于各子目标熵值权重的欧氏贴近度作为适应度值计算方法,引导种群向Pareto前端进化。在进化过程中,运用快速非支配排序策略构建Pareto档案,并利用进化种群中最优个体实时更新Pareto最优解集,提升算法的优化性能;同时通过引入变异算子增加种群多样性,避免算法陷入局部最优。最后,基于Benchmark算例的仿真实验,验证了该算法求解多目标作业车间调度问题的有效性。     

蛙跳优化算法求解多目标无等待流水线调度

提出了基于Pareto边界和档案集的改进蛙跳算法,解决以最大完工时间、最大拖后时间和总流经时间为目标值的无等待流水线调度问题．首先,采用NEH（Nawaz—Enscore—Ham）启发式与随机解相结合的初始化方法,保证了初始群体的质量和分布性;其次,采用两点交叉方法生成新解,使蛙跳算法能够直接用于解决调度问题;再次,利用非支配解集动态更新群体,改善了群体的质量和多样性;最后,将基于插入邻域的快速局部搜索算法嵌入到蛙跳算法中,增强了算法的开发能力和效率．仿真试验表明了所得蛙跳算法的有效性和高效性．     

冷轧机组批量作业计划模型与算法

针对编制冷轧机组作业计划受到钢卷宽度跳跃、入口厚度跳跃和出口厚度跳跃等多个工艺约束的问题, 把排产过程归纳为非对称双旅行商问题, 建立了冷轧机组生产作业计划的Pareto多目标模型. 提出了基于Pareto非支配集的自适应多目标蚁群算法, 利用自适应蚁群算法和Pareto非支配集思想, 综合考虑多个目标, 自适应地提供蚂蚁路径搜索参数, 并对得到的非支配解集对应路径更新信息素, 引导蚂蚁向最优解集方向搜索, 最终提供多个可行的批量作业计划, 根据生产要求从中选择合适的最优排产结果. 利用某冷轧薄板厂实际的生产数据进行仿真实验, 表明模型与算法在冷轧机组批量作业计划编制过程中具有可行性.     

Hybrid sliding level Taguchi-based particle swarm optimization for flowshop scheduling problems

A hybrid sliding level Taguchi-based particle swarm optimization (HSLTPSO) algorithm is proposed for solving multi-objective flowshop scheduling problems (FSPs). The proposed HSLTPSO integrates particle swarm optimization, sliding level Taguchi-based crossover, and elitist preservation strategy. The novel contribution of the proposed HSLTPSO is the use of a PSO to explore the optimal feasible region in macro-space, the use of a systematic reasoning mechanism of the sliding level Taguchi-based crossover to exploit the better solution in micro-space, and the use of the elitist preservation strategy to retain the best particles of multi-objective population for next iteration. The sliding level Taguchi-based crossover is embedded in the PSO to find the best solutions and consequently enhance the PSO. Using the systematic reasoning way of the Taguchi-based crossover with considering the influence of tuning factors α, β and γ is presented in this study to solve the conflicting problem of non-feasible solutions and to find the better particles. As a result, it exhibits a significant improvement in Pareto best solutions of the FSP. By combining the advantages of exploration and exploitation, from the computational experiments of the six test problems, the HSLTPSO provides better results compared to the existing methods reported in the literature when solving multi-objective FSPs. Therefore, the HSLTPSO is an effective approach in solving multi-objective FSPs.     

一种并行多目标遗传邻域搜索算法

现有的多目标遗传算法在解决大规模多目标生产调度问题时虽然有效,但往往非常耗时,难以应用于实际.为了提高求解效率,提出了一种并行多目标遗传邻域搜索算法来求解Pareto边界.该算法将多目标遗传算法的进化方向划分为若干范围,然后同时对每个进化方向的范围使用多目标遗传邻域搜索算法,并行地搜索各方向范围内的Pareto边界;在各进化方向范围内进化的子种群会定期交流各自进化成果.多目标遗传邻域搜索算法的并行化在不增加求解时间的前提下,提高了求解精度,加快了算法的收敛速度.仿真实验结果验证了算法的可行性与有效性.     

A machine-learning based memetic algorithm for the multi-objective permutation flowshop scheduling problem

In recent years, the historical data during the search process of evolutionary algorithms has received increasing attention from many researchers, and some hybrid evolutionary algorithms with machine-learning have been proposed. However, the majority of the literature is centered on continuous problems with a single optimization objective. There are still a lot of problems to be handled for multi-objective combinatorial optimization problems. Therefore, this paper proposes a machine-learning based multi-objective memetic algorithm (ML-MOMA) for the discrete permutation flowshop scheduling problem. There are two main features in the proposed ML-MOMA. First, each solution is assigned with an individual archive to store the non-dominated solutions found by it and based on these individual archives a new population update method is presented. Second, an adaptive multi-objective local search is developed, in which the analysis of historical data accumulated during the search process is used to adaptively determine which non-dominated solutions should be selected for local search and how the local search should be applied. Computational results based on benchmark problems show that the cooperation of the above two features can help to achieve a balance between evolutionary global search and local search. In addition, many of the best known Pareto fronts for these benchmark problems in the literature can be improved by the proposed ML-MOMA.     

An efficient metamodel-based multi-objective multidisciplinary design optimization framework

This paper presents an efficient metamodel-based multi-objective multidisciplinary design optimization (MDO) architecture for solving multi-objective high fidelity MDO problems. One of the important features of the proposed method is the development of an efficient surrogate model-based multi-objective particle swarm optimization (EMOPSO) algorithm, which is integrated with a computationally efficient metamodel-based MDO architecture. The proposed EMOPSO algorithm is based on sorted Pareto front crowding distance, utilizing star topology. In addition, a constraint-handling mechanism in non-domination appointment and fuzzy logic is also introduced to overcome feasibility complexity and rapid identification of optimum design point on the Pareto front. The proposed algorithm is implemented on a metamodel-based collaborative optimization architecture. The proposed method is evaluated and compared with existing multi-objective optimization algorithms such as multi-objective particle swarm optimization (MOPSO) and non-dominated sorting genetic algorithm II (NSGA-II), using a number of well-known benchmark problems. One of the important results observed is that the proposed EMOPSO algorithm provides high diversity with fast convergence speed as compared to other algorithms. The proposed method is also applied to a multi-objective collaborative optimization of unmanned aerial vehicle wing based on high fidelity models involving structures and aerodynamics disciplines. The results obtained show that the proposed method provides an effective way of solving multi-objective multidisciplinary design optimization problem using high fidelity models.     

一种改进的多目标约束优化差分进化算法

提出一种新的多目标优化差分进化算法用于求解约束优化问题.该算法利用佳点集方法初始化个体以维持种群的多样性.将约束优化问题转化为两个目标的多目标优化问题.基于Pareto支配关系,将种群分为Pareto子集和Non-Pareto子集,结合差分进化算法两种不同变异策略的特点,对Non-Pareto子集和Pareto子集分别采用DE/best/1变异策略和DE/rand/1变异策略.数值实验结果表明该算法具有较好的寻优效果.