Research on the Approach of Dynamic Collection and Feature-Based Ship-to-Shore Transmission of Marine Equipment Operation and Maintenance Data Based on Deep Learning

The traditional operational maintenance (O&M) tasks for marine equipment involve the continuous collection of operational status data and full data transmission between the ship and shore. While this method provides comprehensive monitoring information, it often incurs significant data monitoring, storage, and transmission costs. These issues are particularly pronounced in marine environments where bandwidth is limited or the power supply for monitoring devices is constrained. This study proposes a Deep Learning-based method for Dynamic Collection of operational maintenance data and Transmission of data features between the Ship and Shore, termed the DLDCTSS. By employing a deep learning anomaly detection model, DLDCTSS establishes a closed-loop feedback mechanism for dynamic sampling. Specifically, vibration sensors are activated for data collection only when abnormalities are detected, and critical state feature data are selectively transmitted based on O&M requirements. A theoretical analysis demonstrates that the DLDCTSS approach effectively reduces onboard storage overhead, lowers communication energy consumption between ship and shore, and improves data transmission efficiency, cutting overall system expenses. This study first assesses various anomaly detection models on open-source datasets, evaluating their suitability in maritime contexts. Subsequently, tests on a custom water-lubricated stern bearing platform validate both the anomaly detection model and the DLDCTSS approach. This dynamic sampling strategy not only diminishes redundant data collection but also ensures vital information is captured at critical moments, maximizing monitoring quality while enhancing operational efficiency. Meanwhile, transmitting only essential feature data markedly lowers bandwidth usage, and onboard feature extraction safeguards data privacy and security, meeting shipowners’ requirements.