Ocean front reconstruction method based on K-means algorithm iterative hierarchical clustering of sound velocity profiles

Abstract

The invention belongs to the fields of physical ocean, ocean technology, underwater acoustic engineering and the like, and particularly relates to an ocean front reconstruction method based on K-means algorithm iterative hierarchical clustering of sound velocity profiles, and the method uses the K-means algorithm to perform hierarchical clustering on the sound velocity profiles of an ocean front sea area according to a hierarchical principle. Compared with other clustering algorithms, the method is simple in thought, high in convergence speed, good in clustering effect and capable of meeting the clustering requirement when a large number of sound velocity profiles exist in the ocean front sea area. And the accuracy of the clustering result can be verified through the calculation of the sound propagation loss, and the iterative optimization is provided for the hierarchy according to the error threshold until all the hierarchical clustering results meet an allowable error threshold, and finally the ocean front reconstruction based on K-means algorithm iterative hierarchical clustering of sound velocity profiles is realized.

Description

technical field [0001] The invention belongs to the fields of physical ocean, ocean technology, and hydroacoustic engineering, and specifically relates to an ocean front reconstruction method based on K-means algorithm iterative layered clustering sound velocity profile, which is suitable for ocean front reconstruction using temperature, salt, and depth data in the entire sea area. Geometry reconstruction. Background technique [0002] As one of the common mesoscale phenomena in the ocean, the ocean front is defined as a narrow transition zone between two water masses with obviously different properties. The sound velocity profile on both sides of the front is very different. For a strong front, the change of sound velocity will be more than 30m / s, while for a weaker front, the change of sound velocity will be less than 5m / s. The detection of ocean fronts using cluster analysis of sound velocity profiles is a new method. As far as current research is concerned, many studie...

AI Technical Summary

Problems solved by technology

[0003] (1) Using cluster analysis to classify the sound velocity profiles of the studied sea area, the studied sea area is relatively large and includes a variety of ocean phenomena, but there is a lack of targeted research on the mesoscale phenomenon of the ocean front

[0004] (2) Clustering analysis is aimed at the full-depth sound velocity profile, and the hierarchical clustering of the sound velocity profile cannot be realized, and there is no definite division principle for the number of layers and the depth of the layer, and the obtained reconstruction results of the ocean front cannot objectively reflect its geometry. Variation of structure with depth

[0005] (3) The results obtained by clustering cannot be reasonably verified

Images