Research On Polar Marine Environment Information Fusion Algorithm And Database Construction Technology

In recent years,the continuous reduction of the Arctic ice and snow area has gradually shown that some waters of the Arctic Ocean are ice-free in summer,and commercial navigation in the Arctic has become a reality.The Arctic Waterway provides a convenient transportation route for trade between my country and Europe and North America,and it is also an important supplementary flight segment of my country's "One Belt,One Road".Marine environmental information(environmental elements such as sea breeze,waves,currents,etc.)is an important factor affecting shipping safety.Accurate and sufficient marine environmental information can analyze sea conditions for ships and designate recommended routes to assist ships in safe and efficient navigation.At this stage,the polar marine environment information is scarce,and there are still missing phenomena in the few existing polar marine environmental data,which makes it impossible to fully analyze the changes in polar marine environmental information,and thus cannot efficiently serve the polar shipping Safety.Therefore,it is very important to obtain high-quality polar marine environmental data and improve the availability of existing data.The thesis first analyzes the existing marine environmental information sharing platform,and determines the marine environmental information elements that affect the ship's navigation in the polar regions.In order to obtain high-quality polar region marine environmental information from various shared platforms,in view of the uneven quality of marine environmental information data of each platform,combined with the availability evaluation index system of big data and spatial data,establish a system suitable for polar marine environmental information A two-level usability index evaluation system is used to obtain high-quality polar marine environmental information data from various platforms.Secondly,the paper aims at the lack of marine environmental information in polar regions caused by the interference of observation equipment by environmental factors or the instability of acquisition equipment.According to the characteristics of the lack of marine environmental information data in polar regions,a convolutional neural unit-based design is designed.Generative Adversarial Network(GAN)filling algorithm,the generator and discriminator of the algorithm both use Convolutional Neural Unit(CNN).In order to improve the accuracy of data filling,the convolutional network is set to multi-channel to achieve the purpose of fusing various marine environmental information.The Adam optimization algorithm and Dropout are used in the training to cross-train the network to improve the quality of the generated data,so that the generated data samples are used to fill in the missing values in the original data.Thirdly,in view of the problem that the convolutional neural unit-based generative confrontation network filling algorithm cannot correlate the potential relationship between adjacent data points,the GAN polar ocean environment information filling algorithm based on the gated recurrent unit(GRU)is designed.The attenuation mechanism is introduced in the control unit to determine the proportion of historical memory vectors,so that the algorithm can fully explore the potential relationship between missing data and non-missing data,and improve the repair ability of the network.The experimental results under the measured data of polar marine environment information show that the algorithm can effectively capture the nonlinear dynamic characteristics of the polar marine environmental element data and mine the potential associations between the data.Has a higher filling accuracy.Finally,aiming at the problem of the lack of service pertinence of marine environmental information in the polar regions,the paper studies the structure of the S-57 electronic chart data model,and realizes the analysis and display of S-57 electronic chart data.In addition,in view of the redundancy problem of stored data,an improved data relationship model of polar marine environment information is proposed.Then it is developed based on QT and My SQL platform to realize the query of the marine environment information elements of the specified latitude and longitude.