Uma metodologia para avaliaÃÃo da confiabilidade humana em atividades de substituiÃÃo de cadeias de isoladores em linhas de transmissÃo

AUTOR(ES)
DATA DE PUBLICAÇÃO

2005

RESUMO

Human Reliability Analysis (HRA) has been used to study the execution of human actions, and their interactions with a system, taking into account its limitations and factors influencing the human performance. The HRA models are usually categorized into the first and second generation methods, where most of then has been developed in the context of the nuclear industry. The first generation methods present some shortcomings such as unreal assumptions of independence and simple binary representation of events, which mainly result from the use of event and fault trees. Therefore, they present a significant deficiency in modeling human actions as well as quantifying the causal models. Therefore, it is clear that modeling of causalities in human actions has become a great challenge for HRA. In this context, in this dissertation it is described a methodology for modeling human actions through Bayesian belief networks (BBN). The methodology provides a greater flexibility as not only allows for a more realistic representation of the dynamic nature of man-system and man-man interface under normal and abnormal process states, but also represents the relationship of dependence among the events and performance shaping factors. This dissertation uses Bayesian networks for evaluating human reliability in activities of substitution of the insulatorsâ chain in transmission line. It clearly models the human actions, as well as the methods used to build the network, with emphasis in the Bayesian networksâ mechanisms of quantification

ASSUNTO(S)

performance shaping factors engenharia de producao expert judgments method for knowledge elicitation redes bayesianas bayesian inference mÃtodo de eduÃÃo do conhecimento opiniÃo de especialistas fatores de desempenho human reliability analysis (hra) anÃlise de confiabilidade humana (ach) bayesian networks inferÃncia bayesiana

Documentos Relacionados