期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2017
卷号：17
期号：2
页码：9-19
出版社：International Journal of Computer Science and Network Security
摘要：In this paper, evaluation of errors detected in the fluid transmission lines by magnetic PIG in the presence of external magnetic fields, using Fuzzy Inference System, FIS, in order for classification of deterioration and the diagnosis of the deterioration type of structural defects of steel pipes in the petrochemical industry has been done. The set of features mentioned in this article are: 1) geometric features extracted from the raw data of the smart magnetic PIG, 2) features resulting from the response of the model to the current situation. Initially, the strategy of testing is defined and then the required data are collected using MATLAB software. Then, with the help, a parametric transfer function for each pulse is obtained. When the data are considered as the input of the function, it will also be used as the output. 3 selected modes of pipes in this paper are: healthy state, worn and defected. Therefore, the defected state refers to synthetic metals or any other defects. Then, the required characteristics are calculated and they are used in the Fuzzy Inference System as input in the classification section. The obtained system has high capability for classification and detection of deterioration of the pipes with minimum error alarm and low dissatisfaction.
关键词：Detected error; fluid transmission pipeline; PIG; magnetic fields
