Abstract
Marine engine operating regions under principal component analysis (PCA) to evaluate ship performance and navigation behavior are presented in this study. A data set with ship performance and navigation information (i.e. a selected vessel) is considered to identify its hidden structure with respect to a selected operating region of the marine engine. Firstly, the data set is classified with respect to the engine operating points (i.e. operating modes), identifying three operating regions for the main engine. Secondly, one engine operating region (i.e. a data cluster) is analyzed to calculate the respective principal components (PCs). These PCs represent various relationships among ship performance and navigation parameters of the vessel and those relationships with respect to the marine engine operating region are used to evaluate ship performance and navigation behavior. Furthermore, such knowledge (i.e. PCs and parameter behavior) can also be used for sensor fault identification and data compression/expansion types of applications as a big data solution in shipping.