ارزیابی ریسک برای خطوط لوله با نقص فعال بر اساس روش های هوش مصنوعی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|52436||2009||9 صفحه PDF||سفارش دهید||5714 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : International Journal of Pressure Vessels and Piping, Volume 86, Issue 7, July 2009, Pages 403–411
The paper provides another insight into the pipeline risk assessment for in-service pressure piping containing defects. Beside of the traditional analytical approximation methods or sampling-based methods safety index and failure probability of pressure piping containing defects will be obtained based on a novel type of support vector machine developed in a minimax manner. The safety index or failure probability is carried out based on a binary classification approach. The procedure named classification reliability procedure, involving a link between artificial intelligence and reliability methods was developed as a user-friendly computer program in MATLAB language. To reveal the capacity of the proposed procedure two comparative numerical examples replicating a previous related work and predicting the failure probabilities of pressured pipeline with defects were presented.