Intelligent full-stage stable fault diagnosis method for subsea production system

Abstract

Subsea production system is a large slowly changing system. Through fault diagnosis, the state of it can be got in time, and this can ensure the smooth production of oil and gas. When the training data is insufficient, the performance of data driven methods is poor. Influenced by slow changes, the performance of model-based fault diagnosis methods may fluctuate. For long-term stable monitoring of subsea production system, an intelligent full-stage stable fault diagnosis method for subsea production system is proposed. A fault diagnosis framework with both digital and model drives is proposed. Model-based diagnostic model is used to achieve accurate diagnosis under insufficient data. Data driven model is trained continuously to achieve stable diagnosis. A dynamic digital twin model is proposed to accelerate the process of training. A case from the South China Sea is used to study the performance of this method. The results show that this method is stable over a long period.