Iterative beam search algorithms for the permutation flowshop

Libralesso, Luc; Focke, Pablo Andres; Secardin, Aurélien; Jost, Vincent

doi:10.1016/j.ejor.2021.10.015

Cited by 11 publications

(6 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We implemented a single-shot version of the iterative widening beam search approach by Libralesso et al (2021) for the flowtime variant of the PFSP within our framework. The achieved results over the famous Taillard benchmark (Taillard 1993) are shown to be competitive (Libralesso et al 2021) with their introduced guidance which combines a layer-dependent weighted sum of the machines' idle times and the costs-so-far.…”

Section: Preliminary Resultsmentioning

confidence: 99%

“…Beam search has been used to construct high-quality feasible solutions in the context of branch-and-bound, standalone, or combined in a hybrid setting with an improvement heuristic like local search. Quite recently, strong results have been obtained on difficult scheduling problems, see, e.g., Libralesso et al (2021) on Permutation Flow Shop Scheduling (PFSP) and Frohner, Neumann, and Raidl (2020) on the Traveling Tournament Problem (TTP). The most crucial parts are always the evaluation of the nodes, the guidance, and the beam width, limiting the maximum width of the search tree.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Parallel Beam Search for Combinatorial Optimization (Extended Abstract)

Frohner¹,

Gmys

Melab

et al. 2022

SOCS

View full text Add to dashboard Cite

Inspired by the recent success of parallelized exact methods to solve difficult scheduling problems, we present preliminary results of a general parallel beam search framework for combinatorial optimization problems. Beam search is a constructive metaheuristic traversing a search tree layer by layer while keeping in each layer a bounded number of promising nodes to consider many partial solutions in parallel. We propose a variant which is suitable for intra-node parallelization by multithreading with data parallelism. For sufficiently large problem instances and beam widths our work-in-progress implementation in the JIT-compiled Julia language admits promising speed-ups over 30x on 32 cores with uniform memory access for the Permutation Flow Shop Scheduling (PFSP) problem with flowtime objective.

show abstract

Section: Preliminary Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Parallel Beam Search for Combinatorial Optimization (Extended Abstract)

Frohner¹,

Gmys

Melab

et al. 2022

SOCS

View full text Add to dashboard Cite

show abstract

“…The behavior of this approach is biased by arranging jobs the first-row of the matrix according to the starting solution. Moreover, for each visited subproblem of a predefined depth, the beam-search algorithm recently proposed by (Libralesso et al 2020) is applied on partial solutions, leaving prefix and postfix partial schedules unchanged.…”

Section: Heuristic Threadsmentioning

confidence: 99%

“…Due to the large number of open instances in the VRF benchmark (271/480), no attempts are made to solve instances with m = 60 machines and the allocated computational budget per instance is smaller than for the Taillard benchmark. Overall, 55 instances are solved exactly for the first time and 122 best-known solutions (Vallada et al 2015, Libralesso et al 2020, Kizilay et al 2019, Gmys et al 2020) are improved. The largest exactly solved instance is VRF30 20 1 (53.4 × 10 12 decomposed nodes, 200 GPUh)-the previously best-known solution is optimal.…”

Section: Vrf Instancesmentioning

confidence: 99%

Exactly Solving Hard Permutation Flowshop Scheduling Problems on Peta-Scale GPU-Accelerated Supercomputers

Gmys

2022

INFORMS Journal on Computing

View full text Add to dashboard Cite

Makespan minimization in permutation flow-shop scheduling is a well-known hard combinatorial optimization problem. Among the 120 standard benchmark instances proposed by E. Taillard in 1993, 23 have remained unsolved for almost three decades. In this paper, we present our attempts to solve these instances to optimality using parallel Branch-and-Bound (BB) on the GPU-accelerated Jean Zay supercomputer. We report the exact solution of 11 previously unsolved problem instances and improved upper bounds for eight instances. The solution of these problems requires both algorithmic improvements and leveraging the computing power of peta-scale high-performance computing platforms. The challenge consists in efficiently performing parallel depth-first traversal of a highly irregular and fine-grained search tree on distributed systems composed of hundreds of massively parallel accelerator devices and multicore processors. We present and discuss the design and implementation of our permutation-based BB and experimentally evaluate its parallel performance on up to 384 V100 GPUs (2 million CUDA cores) and 3840 CPU cores. The optimality proof for the largest solved instance requires about 64 CPU-years of computation—using 256 GPUs and over 4 million parallel search agents, the traversal of the search tree is completed in 13 hours, exploring [Formula: see text] nodes.

show abstract

“…The TSP is a problem with numerous relevant applications, and as a result, several studies have been carried out to solve it [42][43][44][45]. The SOP, like the TSP, also has several important studies, including energy optimization in compilers [46], search optimization [39,47] and parallel machine scaling [48]. Moreover, the TSP and SOP are NP-hard combinatorial optimization problems; that is, in practice, it is necessary to adopt approximate algorithms in the search for better solutions [39].…”

Section: Introductionmentioning

confidence: 99%

Transfer Reinforcement Learning for Combinatorial Optimization Problems

Souza,

Santos,

Ottoni

et al. 2024

Algorithms

View full text Add to dashboard Cite

Reinforcement learning is an important technique in various fields, particularly in automated machine learning for reinforcement learning (AutoRL). The integration of transfer learning (TL) with AutoRL in combinatorial optimization is an area that requires further research. This paper employs both AutoRL and TL to effectively tackle combinatorial optimization challenges, specifically the asymmetric traveling salesman problem (ATSP) and the sequential ordering problem (SOP). A statistical analysis was conducted to assess the impact of TL on the aforementioned problems. Furthermore, the Auto_TL_RL algorithm was introduced as a novel contribution, combining the AutoRL and TL methodologies. Empirical findings strongly support the effectiveness of this integration, resulting in solutions that were significantly more efficient than conventional techniques, with an 85.7% improvement in the preliminary analysis results. Additionally, the computational time was reduced in 13 instances (i.e., in 92.8% of the simulated problems). The TL-integrated model outperformed the optimal benchmarks, demonstrating its superior convergence. The Auto_TL_RL algorithm design allows for smooth transitions between the ATSP and SOP domains. In a comprehensive evaluation, Auto_TL_RL significantly outperformed traditional methodologies in 78% of the instances analyzed.

show abstract

Iterative beam search algorithms for the permutation flowshop

Cited by 11 publications

References 38 publications

Parallel Beam Search for Combinatorial Optimization (Extended Abstract)

Parallel Beam Search for Combinatorial Optimization (Extended Abstract)

Exactly Solving Hard Permutation Flowshop Scheduling Problems on Peta-Scale GPU-Accelerated Supercomputers

Transfer Reinforcement Learning for Combinatorial Optimization Problems

Contact Info

Product

Resources

About