Modelos alternativos em filas M/G/1

• Ano de defesa: 2015
• Autor(a) principal: Prado, Silvia Maria
• Orientador(a): Louzada Neto, Francisco
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Estatística - PPGEs
• Palavras-chave em Português: Modelos de fila M/G/l; Distribuição Conwav-Maxwell-Poisson; Distribuição do máximo e do mínimo tempos de serviço
• Palavras-chave em Inglês: M/G/l queue; Conwav-Maxwell-Poisson distribution; Distribution for the maximum and minimum service time
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::PROBABILIDADE E ESTATISTICA
• Link de acesso: https://repositorio.ufscar.br/handle/ufscar/8693

The main aim of this work is to develop alternative queuing models to M/ G/l, in which arrivals follow a Poisson process, the total number of customers on the system and the total number of service channels are unknown. Our interest is just to observe the service channel that will offer the maximum or minimum service time. Wherefore, the service distributions are obtained from the composition of the Conwav-Maxwell-Poisson distribution truncated at zero, used to model the number of service channels, with the general distribution to the maximum and minimum service time. Thus, we obtain new distributions for service time, which are called Maximum-Conwav-Maxwell-Poisson-general, denoted by MAXCOMPG distribution, and Minimum-Conwav-Maxwell-Poisson-general, denoted by MINCOMPG distribution, consequently, we obtain the queue models M/MAXCOMPG/1 and M/MINCOMPG/ 1, respectively. As general distributions, we use the distributions exponential, Weibull and Birnbaum Saunders, To illustrate the proposed queue models, a simulation study is done and also real data are used.