Utilização de logica nebulosa na detecção de vazamentos em dutos

AUTOR(ES)
DATA DE PUBLICAÇÃO

2003

RESUMO

Pipeline is an efficient and economic means of transporting petroleum products. However, risks associated with accidental release of transported product are still high. That issue has motivated the development of many methods for leak detection, mainly based on process variables. In the present dissertation, the high correlation between the inlet-outlet flowrate deviation and the operational transients is shown which is the important fact to define the fault detection strategy present here. The applied strategy consists, at first, in a development of classifier module that can identify the operational and process transients and determine the current stage of the transfer process. The output of this module is used by a Fault Detection module that can evaluate the inlet-outlet flowrate deviation in order to detect a leak or a abnormal operation condition, with a low level of spurious alarms. Two fuzzy methods were used to solve this problem: A Fuzzy Inference System using a rule-base developed from this database and a Neural Fuzzy Network using the same role-base. The systems were evaluated with a new data collected from the same process and the results are encouraging with increased leakage or abnormal situation detection, low computational costs and low level of spurious alarms

ASSUNTO(S)

hydrocarbons neural netwoks (computer science) redes neurais (computação) oleodutos fuzzy logic detectores de vazamento logica difusa petroleum pipelenes leak detectors hidrocarbonetos automação automation

ACESSO AO ARTIGO

Documentos Relacionados