Research On Classification Method Of Navigational Aids In Inland Waterway Based On Clustering And Fuzzy Evaluation

Navigational aid, a significant component of modern water transportation, is one of the most important navigation facilities to ensure the safety of vessel, which is designed to play a vital role to help vessel in locating safe position, planning safe route, warning danger and obstacle.The management mode of importance-orientation is playing a more and more significant role in the process of maintaining and managing inland waterway navigational aids, however, definition and classification of key or important navigational aids are not clear, specific and comprehensive, some even rely on personal experience, which is not helpful in the scientific management and maintenance of inland waterway navigational aids. In order to explore a more systematic, scientific and perfect classification method of navigational aids in inland waterway, this thesis proposes a classification method of navigational aids based on data clustering and fuzzy evaluation, which is based on the mass information resources from Telemetering and Remote Control System of Navigational Aids in inland waterway.First of all, some shortcomings existed in the process of maintaining and managing inland waterway navigational aids were analyzed. In the term of the achievements of informationization construction of inland waterway navigational aids management, feasibility of data mining technology applied to the research on classification method of inland waterway navigational aids was discussed. According to the definition of navigational aid efficacy and analysis of the database of Telemetering and Remote Control System of Navigational Aids, five attributes including navigational aid efficacy were selected as the characteristics of classification of inland waterway navigational aids. Additionally, fuzzy comprehensive evaluation model of navigational aids efficacy in inland waterway was constructed for the quantification of navigational aids efficacy.Secondly, data processing of navigational aids was conducted before the improved k-means clustering algorithm (k-means++ algorithm) and hierarchical clustering algorithm applied in the process of navigational aids data clustering. By referring to the idea that clustering before classifying, ranks of navigational aids were divided according to the mean value of each cluster.Finally, based on the designed simulation experiment of classification of navigational aids in inland waterway, the classification results of improved k-means clustering algorithm and hierarchical clustering algorithm were compared and analyzed. Therefore, feasibility and effectiveness of such classification method of navigational aids in inland waterway were proved.