A Negotiation Scenario Using an Agent-Based Modelling Approach to Deal with Dynamic Scheduling

Mezgebe, Tsegay Tesfay; Haouzi, Hind Bril El; Demesure, Guillaume; Pannequin, Rémi; Thomas, André

doi:10.1007/978-3-319-73751-5_29

Cited by 4 publications

(4 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To achieve this, we computed three separate linear layers termed Query (Q), Key (K ), and Value (V ) by mapping of the input vectors p i as shown in (4).…”

Section: Fig 3 Multi-head Attention With Gating Mechanism and Batch N...mentioning

confidence: 99%

“…The majority of job shop scheduling research has focused on producing optimum or good schedules for static environments [2]. In the real world, many production systems face unexpected events that necessitate quick decisions to deal with the changes [3][4][5][6]. Even for a standard job shop scheduling problem, there are several constraints to consider, which considerably complicates the scheduling task.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Gated‐Attention Model with Reinforcement Learning for Solving Dynamic Job Shop Scheduling Problem

Gebreyesus

Fellek

Farid

et al. 2023

IEEJ Transactions Elec Engng

View full text Add to dashboard Cite

Job shop scheduling problem (JSSP) is one of the well‐known NP‐hard combinatorial optimization problems (COPs) that aims to optimize the sequential assignment of finite machines to a set of jobs while adhering to specified problem constraints. Conventional solution approaches which include heuristic dispatching rules and evolutionary algorithms has been largely in use to solve JSSPs. Recently, the use of reinforcement learning (RL) has gained popularity for delivering better solution quality for JSSPs. In this research, we propose an end‐to‐end deep reinforcement learning (DRL) based scheduling model for solving the standard JSSP. Our DRL model uses attention‐based encoder of Transformer network to embed the JSSP environment represented as a disjunctive graph. We introduced Gate mechanism to modulate the flow of learnt features by preventing noise features from propagating across the network to enrich the representations of nodes of the disjunctive graph. In addition, we designed a novel Gate‐based graph pooling mechanism that preferentially constructs the graph embedding. A simple multi‐layer perceptron (MLP) based action selection network is used for sequentially generating optimal schedules. The model is trained using proximal policy optimization (PPO) algorithm which is built on actor critic (AC) framework. Experimental results show that our model outperforms existing heuristics and state of the art DRL based baselines on generated instances and well‐known public test benchmarks. © 2023 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

show abstract

“…To achieve this, we computed three separate linear layers termed Query (Q), Key (K ), and Value (V ) by mapping of the input vectors p i as shown in (4).…”

Section: Fig 3 Multi-head Attention With Gating Mechanism and Batch N...mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Gated‐Attention Model with Reinforcement Learning for Solving Dynamic Job Shop Scheduling Problem

Gebreyesus

Fellek

Farid

et al. 2023

IEEJ Transactions Elec Engng

View full text Add to dashboard Cite

show abstract

“…This mechanism allows to propagate knowledge about perturbations and to return to normal situation in a distributed way without a coordinator. Negotiation approaches are investigated in [13]. They propose a negotiation heuristic based on the notion of critical ratio ((Due_date -current_date) / total shop time remaining).…”

Section: 1mentioning

confidence: 99%

Intelligent Products through SOHOMA Prism

Derigent

McFarlane

El-Haouzi

2021

Service Oriented, Holonic and Multi-Agent Manufacturing Systems for Industry of the Future

View full text Add to dashboard Cite

In the framework of the SOHOMA 2020 special session "SOHOMA 10 th -year anniversary", this paper aims to make a review of the evolution of one important concept studied in the SOHOMA community, namely the Intelligent Product concept. This paper is not review of Intelligent Products -there are several of these already -but rather examines the history of the development of this concept through the 1 st to 9 th editions of SOHOMA, while also proposing future developments on this concept.

show abstract

“…Therefore, exploring the applicability of different MAS negotiation algorithms that help manufacturing components dynamically cooperate for improving performance of the manufacturing system is still a demanding research area. Mezgebe et al (2018) have proposed negotiation model for flexible manufacturing system considering smart product agents scheduled to be processed on different resource agents. The communication protocol was fully controlled by the product agents.…”

Section: Related Workmentioning

confidence: 99%

Multi-agent systems negotiation to deal with dynamic scheduling in disturbed industrial context

et al. 2019

Self Cite

View full text Add to dashboard Cite

It is now accepted that using multi-agent systems (MAS) improve the reactivity to treat perturbation(s) within flexible manufacturing system. Intelligent algorithms shall be used to address these perturbation(s) and all smart decision entities within their environment have to continuously negotiate until their common and final goal is achieved. This paper proposes a negotiation-based control approach to deal with variability on a manufacturing system. It has initially formulated and modeled an environment in which all contributing entities or agents operate, communicate, and interact with each other productively. Then after, simulation and applicability implementation experiments on the basis of full-sized academic experimental platform have been conducted to validate the proposed control approach. Product and resource entities negotiate considering different key performance measures in order to set best priority-based product sequencing. This has been done with expectations that the applicability of the negotiation-based decision-making will be more adaptable to deal with perturbation(s) than another alternative decision-making approach called pure reactive control approach. The result showed that negotiation among the decisional entities has brought significant improvement in reducing makespan and hence conveyed better global performance of a manufacturing system.

show abstract

A Negotiation Scenario Using an Agent-Based Modelling Approach to Deal with Dynamic Scheduling

Cited by 4 publications

References 7 publications

Gated‐Attention Model with Reinforcement Learning for Solving Dynamic Job Shop Scheduling Problem

Gated‐Attention Model with Reinforcement Learning for Solving Dynamic Job Shop Scheduling Problem

Intelligent Products through SOHOMA Prism

Multi-agent systems negotiation to deal with dynamic scheduling in disturbed industrial context

Contact Info

Product

Resources

About