The increasing availability of ship operational data from onboard sensors presents both opportunities and challenges for performance analysis and monitoring. This study proposes a robust ship operational data analysis (SODA) framework to enhance data usability through streamlined processing and analytical procedures. A key component of the framework is a novel metric to quantitatively evaluate the quality of the processed data, which offers a sophisticated tool to compare the performance of different ship operational data analysis strategies, thus is useful for the establishment of optimal data filtering and correction methods. The so-called information entropy weighted data quality index is formulated based on first principles, which systematically evaluates the processed data quality from three aspects: information richness, degree of data scattering, and the correlation between ship speed and engine power. Various filtering strategies and weather influence correction techniques adopted by ISO 15016 to enhance the reliability and consistency of performance metrics derived from the data are systematically investigated. Real-world operational data from a container ship are employed to validate the proposed framework, demonstrating notable improvements in data quality and consistency. The SODA framework can also facilitate the identification of long-term performance trends, highlighting its potential for supporting proactive maintenance planning. Results show that stricter filtering criteria are effective in reducing data scatter and uncertainty, but inevitably suffer from more severe loss of data richness. Milder filtering combined with weather influence corrections allows for more comprehensive trend analysis over extended periods. Overall, the proposed SODA framework provides a practical and robust solution for enhancing ship performance monitoring, enabling data-driven decision-making in maritime operations.
原文链接
分享
引用
