Study On The Fluctuation Law And Forecasting Method Of Coal Waterborne Price In China Coastal Area

In recent years,the fluctuation of coal transport price is more intense,and coastal coal water transport plays an important role in the energy supply in eastern and southern China.The purpose of this paper is to study the fluctuation law of coal water transport price in coastal areas of China in detail and find the most suitable forecasting method,so as to provide reference and suggestions for the government,shipping operators and railway departments.Firstly,this paper clustered the time series data of 10 coastal water transport lines from July 2013 to December 2018 by the system clustering method based on the sum of squares of deviations,and summarized the clustering results.According to the distance and the load of ships on the lines,the coastal coal freight lines in China can be divided into four categories:midway middle tonnage,midway high tonnage,long-distance high tonnage and midway low tonnage.Based on the classification of lines,this paper firstly makes a descriptive analysis of the price fluctuation of each type of lines.Then,the time series data of each type of lines are disassembled by X-12-ARIMA seasonal adjustment method and H-P filtering method respectively.The trend,seasonality,periodicity and irregularity of each type of lines are analyzed in detail.It is found that each type of lines has its own unique fluctuation law.It is concluded that the overall fluctuation law of coastal coal water transport price in China is the trend of first falling and then rising,the seasonality of first rising and then falling,the non-obvious periodicity and the irregularity influenced by policy changes and abnormal weather.After completing the analysis of the fluctuation law of various routes,this paper further explores the most suitable model for predicting the price of coastal coal water transport in China.After comparing the predictive ability of three different models for each category,it is proposed that the X12-HP-NAR combined model method,which combines time series decomposition with neural network prediction,is the most suitable prediction model for this area.Finally,based on the above research,this paper draws the conclusion that specific analysis and combination model are most suitable for short-term forecasting when researching the price of coal cargo in coastal areas of China.The paper also gives suggestions about how to deal with price fluctuations for shipping operators and railway departments respectively.