Robust solutions in multi-objective stochastic permutation flow shop problem

González-Neira, Eliana María; Urrego-Torres, Ana María; Cruz-Riveros, Ana María; Henao-García, Catalina; Montoya‐Torres, Jairo R.; Molina-Sánchez, Lina Paola; Jiménez, Jose-Fernando

doi:10.1016/j.cie.2019.106026

Cited by 22 publications

(9 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To solve the NP-hard problem, many hybrid meta-heuristics have been proposed [ 2 , 3 ]. For the PFSP, its objective is to find a good job sequence to minimize makespan [ 4 , 5 , 6 ], flow time [ 7 , 8 , 9 , 10 ], tardiness [ 11 , 12 , 13 , 14 ], multiple objective [ 15 , 16 , 17 , 18 , 19 , 20 ], etc.…”

Section: Introductionmentioning

confidence: 99%

Intelligent Decision-Making of Scheduling for Dynamic Permutation Flowshop via Deep Reinforcement Learning

Yang

Wang

2021

Sensors

View full text Add to dashboard Cite

Dynamic scheduling problems have been receiving increasing attention in recent years due to their practical implications. To realize real-time and the intelligent decision-making of dynamic scheduling, we studied dynamic permutation flowshop scheduling problem (PFSP) with new job arrival using deep reinforcement learning (DRL). A system architecture for solving dynamic PFSP using DRL is proposed, and the mathematical model to minimize total tardiness cost is established. Additionally, the intelligent scheduling system based on DRL is modeled, with state features, actions, and reward designed. Moreover, the advantage actor-critic (A2C) algorithm is adapted to train the scheduling agent. The learning curve indicates that the scheduling agent learned to generate better solutions efficiently during training. Extensive experiments are carried out to compare the A2C-based scheduling agent with every single action, other DRL algorithms, and meta-heuristics. The results show the well performance of the A2C-based scheduling agent considering solution quality, CPU times, and generalization. Notably, the trained agent generates a scheduling action only in 2.16 ms on average, which is almost instantaneous and can be used for real-time scheduling. Our work can help to build a self-learning, real-time optimizing, and intelligent decision-making scheduling system.

show abstract

Section: Introductionmentioning

confidence: 99%

Intelligent Decision-Making of Scheduling for Dynamic Permutation Flowshop via Deep Reinforcement Learning

Yang

Wang

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…These hybridize a metaheuristic approach with a simulation to obtain solutions for stochastic problems. Simheuristics have been successfully used in vehicle routing problems [11][12][13], inventory routing problems [14], facility location problems [15], and scheduling problems [16][17][18][19].…”

Section: Introductionmentioning

confidence: 99%

A Multicriteria Simheuristic Approach for Solving a Stochastic Permutation Flow Shop Scheduling Problem

2021

Self Cite

View full text Add to dashboard Cite

This paper proposes a hybridized simheuristic approach that couples a greedy randomized adaptive search procedure (GRASP), a Monte Carlo simulation, a Pareto archived evolution strategy (PAES), and an analytic hierarchy process (AHP), in order to solve a multicriteria stochastic permutation flow shop problem with stochastic processing times and stochastic sequence-dependent setup times. For the decisional criteria, the proposed approach considers four objective functions, including two quantitative and two qualitative criteria. While the expected value and the standard deviation of the earliness/tardiness of jobs are included in the quantitative criteria to address a robust solution in a just-in-time environment, this approach also includes a qualitative assessment of the product and customer importance in order to appraise a weighted priority for each job. An experimental design was carried out in several study instances of the flow shop problem to test the effects of the processing times and sequence-dependent setup times, obtained through lognormal and uniform probability distributions with three levels of coefficients of variation, settled as 0.3, 0.4, and 0.5. The results show that both probability distributions and coefficients of variation have a significant effect on the four decision criteria selected. In addition, the analytical hierarchical process makes it possible to choose the best sequence exhibited by the Pareto frontier that adjusts more adequately to the decision-makers’ objectives.

show abstract

“…In the experiments, they analyze the probability of accomplishing the deadlines of a given solution. González-Neira et al (2019) developed a Tabu Search simheuristic to solve a stochastic PFSP. They tested two different probability distributions (Log-normal and uniform) for the processing times with three levels of uncertainty, analyzing solutions according to the expected tardiness and standard deviation.…”

Section: Simheuristicsmentioning

confidence: 99%

“…Therefore, instead of considering deterministic processing time p i for each operation O i ∈ O, a random variable P i with E[P i ] = p i is used to model stochastic processing times for the operations in order to better deal with the uncertain environment. Based on previous works in the scheduling literature (Juan et al 2014, González-Neira et al 2019, Hatami et al 2018, Villarinho et al 2021, we use Log-Normal distributions for modeling stochastic processing times. The Log-Normal distribution has two parameters, µ i and by the expressions (3-1) and (3-2), respectively (Juan et al 2011):…”

Section: Stochastic Plsvspmentioning

confidence: 99%

“…On the other hand, a high number of replications might result in computational overhead for the method. To overcome these issues, we incorporate a strategy proposed by González-Neira et al (2019) of using the confidence interval's error around the mean, from the available observations, to control the number of simulation replications. Thus, after each replication, confidence intervals at a given confidence level are calculated, and the algorithm computes the interval error.…”

Section: Iterated Greedy Simheuristic With Embedded Monte Carlo Simulation For the Stochastic Plsvspmentioning

confidence: 99%

See 1 more Smart Citation

Solving the Deterministic and Stochastic Pipe-Laying Support Vessel Scheduling Problem

CUNHA¹

View full text Add to dashboard Cite

show abstract

Robust solutions in multi-objective stochastic permutation flow shop problem

Cited by 22 publications

References 58 publications

Intelligent Decision-Making of Scheduling for Dynamic Permutation Flowshop via Deep Reinforcement Learning

Intelligent Decision-Making of Scheduling for Dynamic Permutation Flowshop via Deep Reinforcement Learning

A Multicriteria Simheuristic Approach for Solving a Stochastic Permutation Flow Shop Scheduling Problem

Solving the Deterministic and Stochastic Pipe-Laying Support Vessel Scheduling Problem

Contact Info

Product

Resources

About