Research On Green Tide Forerast Based On Big Data Technology

Since 2007,green tides have been appearing for the first time in the sea near Yancheng,Jiangsu Province,and their growth,aggregation and reproduction are largely influenced by temperature.as well as precipitation,light and other factors.And the drift of the green tide is generated by the combined action of wind fields and ocean currents.Final demise.The generation and accumulation of green tides have already caused harm to the marine ecological environment and endangered the ecological balance of marine life.China's prediction and early warning of the green tide disaster has become urgent.The study of the distribution area of green tide and the trend of green tide drift path can help to the management of green tide.In response to the complexity of green tide data and the number of influencing factors,and the difficulty of obtaining correlations between data.In this paper,a prediction model based on gray correlation analysis and rough set optimization of BP neural networks is proposed.The model is based on the BP neural network and uses gray correlation analysis to calculate the correlation values of green tide data,filter out the data with low correlation and delete them.Second,introducing an approximately simple algorithm for the importance of properties in the rough set.The properties of the green tide data are approximated by calculating the importance values.Combining dichotomous partitioning with an empirical formula to obtain the number of nodes in the implied layer,which solves the problem of difficulty in determining the number of nodes in the implicit layer for BP neural networks.Finally,based on the green tide data of the Yellow Sea from 2012 and 2014-2018,experimental analysis by MATLAB.The results show that the predictive model in this paper is able to achieve improvements in the speed of network convergence with certain accuracy.Take the 2014-2018 Enteromorpha prolifera remote sensing data as real instance,and based on the Arc GIS technology platform,the concept of the center of gravity of physics is combined with the POM numerical model to study the drifting trend of the green tide.Consider the green tide as a geometric block of uniform mass,without regard to its own factors.The center of gravity of distribution of the green tide represents the concentration point of the force during the drifting motion.Therefore,the coordinates of the center of gravity of the green tide on that day are calculated using the MeanCenter method in spatial analysis to analyze its drift trajectory.Based on this,using the POM numerical ocean model to simulate flow field data to provide input data for green tide drift prediction.The location of the green tide at the next moment is calculated based on the results of the data obtained from the simulation by the Lagrangian particle tracking method to form a drift trend graph.