/system/images/000/000/219/Logo_Journ_es_Optimisation__2012_Outl_default.png

Incluant une Journée industrielle de l'optimisation

HEC Montréal, 7 - 9 mai 2012

JOPT2012

HEC Montréal, 7 — 9 mai 2012

Horaire Auteurs Mon horaire

WA3 Simulation et systèmes à agents / Simulation and Multi-agent Based Systems

9 mai 2012 09h00 – 10h40

Salle: Metro inc.

Présidée par Jean-Marc Frayret

4 présentations

  • 09h00 - 09h25

    Optimization of Sorting and Allocation Activities in Sawmills, by using Data-Mining Techniques and Agent-Based Modeling

    • Alvaro Gil, prés., Polytechnique Montréal

    In this presentation, we will propose an agent-based model of log yard activities to study how the processes of sorting and allocating log classes to customer orders affect lumber production yields. This model uses simple classification rules, generated with data mining techniques and classification trees. These rules are tested and analyzed using both agent-based and discrete event simulations.

  • 09h25 - 09h50

    Improving Quality of Service in the Physical Internet

    • Chun Wang, Concordia University
    • Bhuiyan Fozlul, prés., Université Concordia

    Quality of Service (QoS) is an important concern for sustainable transportation system. We propose an auction-based approach for improving quality of service in the Physical Internet environment, where logistic activities like handling physical objects, sorting, routing, and storage are coordinated among service providers located at a large number of nodes within the network.

  • 09h50 - 10h15

    A Research of Computer Network Systems' Productivity and Efficiency with Simulation Modeling

    • Sultan Ceren Salkın, prés., Isfahan University of Technology

    Computer networks are one of the most prevalent systems with many applications that we use in our daily lives. Almost all companies, factories and internet sale organizations utilize network systems to respond customer demands. Although network systems seem simple, actually they are not. Behind the system, links are working with different complex algorithms. The problems with these commands sometimes cause communication troubles and affect most of the job processes. With this study, it is aimed to investigate the efficient and effective working ways of client-server structure in a network topology, to carry out productivity studies and decide what kind of network topology will be convenient for different systems with simulation modeling. First of all, the working principle of the web system was explained, deeply analyzed, the former data about different buttons was gathered from an e- business web system. Secondly, data were suited with a known probability distribution. Thereafter, system situation variables, system criterias, system attributes, performance variables, parameters and processes were all defined. Afterwards, ARENA simulation model was effectuated by entering entity and assigning values for attributes. Additionally, processes and conditional situations are assigned with their properties. Finally, run parameters were entered, and simulation model was run. According to the results, existing system resources usage ratios, queues, performance criteria situations and model verification were criticized and different scenarios were discussed to make suggestions for the improvement of network system. To sum up, it is proposed to show the usage efficiency and the productivity of the computer network systems for making improvement of the system by putting forward the system topology’s work with ARENA simulation model. Hopefully, this study will be a good example for similar network systems.

  • 10h15 - 10h40

    A Simulation Approach of Strategy Analysis of Repeated Timber Auction Allocation

    • Farnoush Farnia, prés., École Polytechnique de Montréal
    • Jean-Marc Frayret, Polytechnique Montréal, CIRRELT
    • Luc LeBel, Université Laval
    • Catherine Beaudry, École Polytechnique de Montréal

    The auction design has become remarkably important in many markets especially natural resources. Using agent based simulation, we model and analyze repeated timber auctions. We simulate and compare different kind of approaches for bidders’ bidding strategies in various setup configurations. The comparisons include the success rate of winning the auction and the winning
    price per m3.

Retour