Wyniki wyszukiwania - Biblioteka Nauki

1

Analiza wydajności agentów podejmujących decyzje kupna–sprzedaży w systemie wieloagentowym a-Trader

100%

Korczak J. , Hernes M. , Bac M.

Informatyka Ekonomiczna

|

2014

|

nr 1(31)

77-90

EN

The article presents the performance analisys issues of buy-sell decisions agents’ in a-Trader system. The system allows for supporting of investment decision on FOREX market. The first part of the article contains a description of a-Trader system. Next, the algorithms of the selected buy-sell decision agents are presented. The evaluation function of agents’ performance is elaborated, and the manner of performance analisys executing is presented in the final part of the article.

2

Alliance and Coalition Formation

75%

Fišer J. , Mashkov V.

|

2010

|

tom Vol. 18, nr 1

19-38

EN

Rapid development of agent technologies has evoked new research problems, among them formation of the groups of agents, such as alliances and coalitions. In the paper we differentiate between alliances and coalitions and suggest approaches to their formation. Agents of the highlighted alliances and coalitions are considered as group-rational, i.e., more interested in achieving the common goal than in their own benefits. The paper deals with specific alliances in which an agent accepts the offer to join the alliance despite the fact that one or more clauses of the planned alliance are not acceptable for him. This way, the agent may participate in the alliance on restricted conditions. The paper focuses on the problems of alliance and coalition formation, specifically, on the issues of agents’ activities at the stage of alliance and coalition structuring. For different cases when an agent is either informed about other agents’ capabilities or not, different agent’s activities are considered. For the purposes of modeling and verification of the proposed algorithms and approaches to alliance and coalition formation the Erlang/OTP platform is used. The paper specifies how the Erlang/OTP platform can be used for modeling agents’ intercommunication during formation of coalitions.

3

Symulacja "wyłaniającej się" samoorganizacji w wieloagentowym wydziale produkcji

63%

Schafer K. , Partiti V. , Charpentier P. , Martin P. , Spath D.

|

1998

|

tom R. 3, z. 2/3

135-147

PL

Coraz większe wymogi, jakie muszą spełniać systemy wytwarzania pod względem elastyczności, odporności i wydajności stanowią bodziec do poszukiwania nowatorskich rozwiązań w tej dziedzinie. Monolityczne, nie dające się przekonfigurować, systemy hierarchiczne nie są już w stanie sprostać zmieniającym się wymaganiom rynku. A zatem wydaje się, że zdecentralizowanie sterowania i danych -jak to ma miejsce w systemach wieloagentowych - umożliwi szybkie reagowanie na zmieniające się warunki. Jednak zanim przedsiębiorstwa wymienią tradycyjny system harmonogramowania na bliżej nieznane nowe podejście, takie jak paradygmat wieloagentowy, trzeba dowieść prawdziwości teorii, na której ono się opiera. Tak więc celem obecnych badań jest, po pierwsze, dostarczenie narzędzia symulacji niejednorodnemu wydziałowi produkcji, po drugie, umożliwienie fizycznego rozproszenia od razu po uzyskaniu przekonywających wyników symulacji. Interfejs użytkownika umożliwia skonfigurowanie animowanej symulacji fizycznego zachowania się wydziału produkcji i agentów programowych, współdziałających ze sobą w procesie negocjacji, których wynikiem jest harmonogramowanie przepływu pracy. Przenośność tego szablonu zapewnia język programowania JAVA, a agenci programowi zostali zaprojektowani zgodnie z obowiązującymi standardami (język komunikacyjny KQML, protokół sieci kontraktowej, standard STEP/EXPRESS, architektura BDI), aby uniknąć rozwiązania odosobnionego.

EN

The growing demand for more flexibility, robustness and productivity in manufacturing systems made way for new scientific approaches formerly not known in this domain. Fast changing market situations can no more be coped with by monolithic hierarchical systems that lack from re-configurability. Therefore, the distribution of control and data - as in multi-agent systems - seems to give the answer to the question of reactivity. However, the theory has to be proven before enterprises will trade their classical scheduling system for an un-known approach like the multi-agent paradigm. Thus the aim of this research work is to furnish a simulation tool for a heterogeneous workshop in the first place, which offers in addition the possibility to be physically distributed on the spot in case the simulation output is convincing. The user interface makes accessible both the configuration of the animated simulation of the physical workshop behaviour and the software agents that interact in a negotiation process resulting in the scheduling of the workflow. The portability of this template is assured by the programming language JAVA, whereas the design of the software agents follows existing acknowledged specifications (the KQML communication language, the Contract Net Protocol, the STEP/EXPRESS Standard, the BDI architecture) in order to avoid an isolated solution.

4

Survey of the agent-based approach to intelligent manufacturing

63%

Madejski J.

|

2007

|

tom Vol. 21, nr 1

67-70

EN

Purpose: Summary of the state-of-the-art of the Distributed Artificial Intelligence applied to Intelligent Manufacturing. Main applications are presented along with different technologies applied in these areas. Design/methodology/approach: Intelligent Manufacturing area was split into many segments, which require different approach to intelligent problem solving. Multiagent systems negotiation needs were analysed and cooperation issues in the form of clustering, cloning, and learning were analysed in search for relevant tools. Findings: Detailed review of the approach to development of the agent based Intelligent Manufacturing from the fundamental considerations to the latest hands-on developments. Research limitations/implications: Many presented technologies call for detailed study before they can be implemented in practice. Originality/value: Thorough review of the Distributed Artificial Intelligence approach to current agile manufacturing needs. Key technologies are pointed out along with the main areas in which they can be implemented, and which require further research.