Effects of biofouling in Baltic Sea region: Applying Chow-Liu-tree augmented Naïve Bayes-theorem for analyzing voyage data of the ship

Abstract

The objective of the COMPLETE-project is to prevent spreading of harmful and invasive alien species in Baltic Sea region. Significant proliferation has happened with ballast waters but organisms are also being transported to new areas among underwater hull structures. In addition of alien species spreading issues, biofouling increases vessel hydrodynamic resistance affecting straightforward to fuel consumption and carbon dioxide emissions. At present day, this bio-contamination is deducted by regular cleanings of immersed hull structures during summer seasons. However, selection of cleaning intervals is based on experience. Also, the effect of cleaning is often perceived by crew, but normally there are no measurement-based knowledge on its effect on voyage of unique vessel. Nowadays ship systems provides increasingly data flow that can be stored automatically. Conclusion-making from big data requires appropriate tools specially limiting effects of many system mixers. This article explores usability of Chow-Liu-tree augmented Naïve Bayes method for analyzing voyage data. The advantages of this method are computational efficiency and ability to produce reliable conclusions about causation relationships prevailing in the studied system, even if available data is quite limited.