Construction of Inland River Hybrid Tugboat based on K-means Clustering

In the research of hybrid electric ship, the sailing condition as the basis of each research is very important to the energy consumption test and the formulation of control strategy. Although there are mature conditions construction methods in the automotive field, the typical test conditions with high reliability are still lacking in the Marine field. Firstly, this paper collects more than 1000 hours of Chinese inland tugboat navigation data, and constructs a detailed tugboat navigation database. By pre-processing the data, ten key feature parameters are extracted. With PCA principal component analysis, the data dimension was reduced, and five typical working conditions were obtained through cluster analysis. By synthesizing fragments with the smallest error, a typical sailing condition conforming to the actual operating conditions of tugboats in inland river ports in China was finally established. After comparative analysis of relative errors, the average relative error between the established working conditions and the characteristic parameters of the database was 4.3%. From the error analysis results, the typical sailing conditions constructed in this study were highly similar to the original working conditions in the database. It can reflect the navigation characteristics of inland tugboat more accurately, and provide the basis for the study of the hybrid power system of inland tugboat in China.