| In today’s ever-increasing application scenarios of information technology,marine science has also used many computer science-related technologies in this field.Ocean reanalysis is based on ocean dynamics model,using data assimilation technology,combining ocean observation data with uneven spatial and temporal distribution with grid point data of numerical models,and finally obtaining the multi-temporal-scale variation characteristics of the ocean elements and multi-element physical associations.Marine reanalysis products have important application value in the fields of commercial oceanography,marine scientific research,and climate change.In this era of data explosion,how to speed up the processing speed of visualization has also become a problem that must be solved.This topic proposes a method of combining parallel computing and parallel query collection to speed up the processing of vector data,and designs experiments to verify the acceleration effect.Finally,Design and implementation of visualization software for ocean current data in marine reanalysis products.First of all,research on the method of drawing the ocean vector field visualization,from the streamline generation algorithm to the streamline integral drawing to the dynamic programming of the streamline,how the streamline can be drawn by computer processing data,and some The most suitable streamline selection principle for the drawing area,and the screening data structure and algorithm for the drawing area.The interpolation algorithm is used to improve the data resolution.Secondly,the current status and development trend of multi-core processors and parallel programming frameworks are analyzed.The analysis and comparison of the selection of big data marine vector field visualization frameworks are analyzed and compared.Based on the characteristics of the current mainstream parallel computing frameworks,the selection is suitable for this A parallel computing framework for the subject,and process design based on this framework.Thirdly,in view of the streamline drawing speed of the big data ocean vector field,this paper proposes a method of combining search and collection and parallel computing to improve the drawing speed of the streamline.First,the principle and implementation method of the data structure of the search and collection will be described in detail.Through the drawing of graphics and examples,the use method of the search and collection is described in detail and the scene it should appear in.The principle of GPU acceleration using MPI is studied and the flowchart of the project acceleration is designed.Finally,this topic is introduced in detail.The proposed accelerated visualization method,that is,the method of combining GPU parallel computing and data collection to perform accelerated visualization display,and design experiments to verify.Finally,according to the project requirements of marine reanalysis products,a visualization software was designed,and a visualization system of ocean vector fields was designed and implemented according to the design requirements.After reference to multiple documents and open source code,the vector field data visualization was independently designed and implemented The simulation program and design experiment have compared and analyzed the software display effect and visualization speed,so that the visualization effect and processing speed basically meet the requirements of the global high-resolution reanalysis product ocean current visualization project. |
