Improving degradation prediction models for failure analysis in topside piping: A neuro-fuzzy approach

Abstract

This manuscript focuses on integrating online condition monitoring data directly into the degradation prediction models. This will aid in-service inspection planning in the identification of possible failures in the topside piping equipment of offshore oil and gas (O&G) production and process facilities (P&PFs). The capability of data clustering and data filtration as well as the interpretation of expert knowledge in artificial intelligent (AI) techniques, such as k-means clustering, artificial neural networks and fuzzy inference systems, has been exploited to meet the aforementioned. The k-means clustering is used in the identification of linguistic parameters from condition monitoring data. Moreover, a neural network approach is used to identify the membership function patterns using online condition monitoring data. The proposed neuro-fuzzy system will help inspection planners to recommend accurate thickness measurement locations (TMLs) for reliable inspection planning programs.