Design of a multi-source monitoring scheme and data fusion method for in-service pressure vessels in the coastal environment

Abstract

The coupling effect of high humidity, high salinity, and alternating loads in coastal environments poses complex corrosion-fatigue damage risks to pressure vessels in gas supply systems. The traditional periodic inspection model cannot provide real-time detection of early damage. To address the non-destructive constraints of in-service equipment, this paper proposes a comprehensive monitoring scheme that integrates real-time online monitoring with periodic offline inspections. First, based on a previously constructed failure mode map, the differentiated monitoring requirements for vessel types are clarified. Then, a non-invasive sensor deployment strategy is designed, and an alternative scheme, “surface strain monitoring + digital twin inversion,” is proposed for multi-layer wrapped vessels. Finally, a three-level data fusion framework of “preprocessing – feature fusion – decision support” is constructed, and a pressure-vessel health index (PVHI) calculation model is established to enable health-status classification and early-warning decision output. Validation using historical data shows that key monitoring parameters can effectively capture trends in the evolution of vessel damage. This scheme provides a feasible technical path for state perception of in-service pressure vessels in coastal environments.