Design And Implementation Of Global Iron Ore Trade And Shipping Analysis System

Iron ore trade plays a decisive role in global trade.The annual global trade volume is 1.5 billion metric tons and the trade volume is over 120 billion US dollars.Iron ore trade is one of the barometers of the world economy.The volume of iron ore trade is usually an afterthought value published through customs statistics.However,the advance forecast of iron ore trade volume has more important significance for traders,futures trading participants and loan companies.This paper uses the ship position data to make trade forecast analysis of iron ore,which can accurately predict the arrival tons of iron ore in the future.This paper mainly completes the forecast of iron ore trade volume through the following process:1.The HDDBSCAN(Hadoop based Distributed DBSCAN)algorithm was designed to identify iron ore loading and unloading berths.The amount of ship position data is very large,and the total position data is more than 100 billion lines(5-years cumulative data).Even if it is an iron ore ship,there are about 1 billion lines.The single-process DBSCAN algorithm performs cluster analysis,and the calculation duration is calculated on an annual basis.In order to solve this problem,this paper designs a distributed DBSCAN algorithm based on Hadoop and geographic region partitioning.As we all know,DBSCAN can only cluster data with similar density.However,the density of ship position data depends on the number of base stations in the area.Therefore,the density of ship position data is very uneven,the density of China is the highest,and the density of South Africa and India is very low.In order to solve the problem of uneven density,the low-density class cannot be obtained,and different DBSCAN algorithm parameters are used for different regions.2.The SVF(Statistic based Voyage Forecast)algorithm is designed to predict ship voyages.The voyage algorithm of the existing literature is aimed at the voyages within the port area and has not been able to achieve global voyage prediction.Based on the partitioning algorithm and the posterior probability of the destination port of each segment,the Bayes algorithm is used to continuously calculate the probability of the destination port with the increase of the segment during the navigation process,and finally achieve the purpose of accurate prediction.Taking the above two algorithms as the core,this paper has completed the following system functions:1.Technical architecture design:The technical architecture is divided into page layer,service layer,business logic layer and data layer.2.Functional module design:Based on business needs,the design and implementation of iron ore trade volume forecast,iron ore real-time shipping and other functional modules.3.Other related modules:system security related,such as preventing SQL injection,user information security,etc.