Research on an intelligent monitoring and diagnosis system for fatigue damage in coke drums based on data-driven approaches

Abstract

As critical high-temperature pressure equipment in the petrochemical industry, coke drums operate for extended periods under complex working conditions. Their structural health status directly impacts production safety and operational efficiency. This paper proposes an intelligent fatigue damage monitoring and diagnosis system based on strain data-driven methods. By collecting and analyzing multidimensional data such as strain and temperature in real time, the system rapidly constructs rainflow matrices and nonlinear cumulative damage models to achieve dynamic diagnosis of fatigue damage, health status classification, and remaining useful life prediction. The system integrates sensing, acquisition, transmission, and edge computing modules. Combined with multi-dimensional data acquisition software and an intelligent diagnosis visualization platform developed on the LabVIEW platform, it enables end-to-end intelligent management from data sensing to condition assessment. Engineering applications demonstrate that the system exhibits excellent reliability, stability, and engineering applicability, providing a feasible solution for structural health management of coke drums and other large-scale pressure equipment.