Seasonal Fluctuations of Steam Coal Prices: Fourier Feature Analysis and XGBoost Forecasting

Siqi Duan

doi:10.47297/wspdecWSP2515-797303.20261004

PDF
Export
Share
Collection
Album

Seasonal Fluctuations of Steam Coal Prices: Fourier Feature Analysis and XGBoost Forecasting

Volume 10, Issue 4

Views: 0 Downloads: 0

Seasonal Fluctuations of Steam Coal Prices: Fourier Feature Analysis and XGBoost Forecasting
Seasonal Fluctuations of Steam Coal Prices: Fourier Feature Analysis and XGBoost Forecasting
School of Economics and Management, East China Normal University, Shanghai, 200062, China
Vol. 10, Issue 4, Pages: 15-19(2026)
Online First：29 July 2026，

Published：30 April 2026
DOI：10.47297/wspdecWSP2515-797303.20261004
Accepted：

Scan QR Code

Duan Siqi.Seasonal Fluctuations of Steam Coal Prices: Fourier Feature Analysis and XGBoost Forecasting[J].Development Economics of China,2026,10(04):15-19.
DOI：

Duan Siqi.Seasonal Fluctuations of Steam Coal Prices: Fourier Feature Analysis and XGBoost Forecasting[J].Development Economics of China,2026,10(04):15-19. DOI： 10.47297/wspdecWSP2515-797303.20261004.

摘要

Abstract

The various components of your paper Thermal coal plays a dominant role in China's energy consumption

making it highly significant to analyse its price fluctuations. This paper collects data on the price of 5

500 kcal/kg thermal coal and relevant influencing factors for the period from 2020 to 2025

primarily to address the challenges in forecasting thermal coal prices caused by seasonal fluctuations. During the data processing and feature engineering stages

this paper first ensures that the price series reaches a stationary state

followed by PCA dimension reduction on a series of macroeconomic indicators. Subsequently

Fourier series are incorporated into the seasonal modelling process

using parametric approximation to capture the cycle of the coming year and generate trigonometric features. Building upon the aforementioned work

this paper develops static and rolling forecasting models based on XGBoost. The performance of the two models is compared using out-of-sample testing

and their application scenarios are identified

thereby providing suitable tools for high-frequency decision-making scenarios.

关键词

Keywords

references

Zhang Huan , Jiang Zuobin . Establishment of an ARIMA Model for Chinese Coal Prices and Its Forecast Analysis [J ] . Industrial Technology and Economics , 2007 , 26 ( 7 ); 102 – 105

Kang Xiaodi . ( 2025 ). Analysis of the Causes of Port Coal Price Fluctuations and Research on Inventory Decision-Making (Master's Thesis, North China Electric Power University (Beijing)) .

Liu Jiaxuan . ( 2024 ). Analysis and Forecasting of Coal Prices Based on a Multi-Feature Fusion Network (Master's Thesis, Nanjing University of Information Science and Technology) .

Cong Jiahui , Liu Zhiping & Xu Gang . ( 2023 ). Modelling and Analysis of Coal Price Forecasting Based on Data Dimension Reduction . Research on Coal Economics , 43 ( 02 ), 18 – 24 .

Yang Zhenqian . ( 2024 ). A Study on Coal Price Forecasting Models Based on Feature Engineering . Research on Coal Economics , 44 ( 11 ), 30 – 37 .

Shu Zhengyu , Fang Manqin , Li Huangqiang , He Fei , Yao Qin & Bao Gang . ( 2025 ). Economic Optimisation of Integrated Energy System Dispatch Accounting for Seasonal Carbon Emissions . Smart Power , 53 ( 09 ), 37 – 47 .

Shao Feng , Feng Yu , Lu Yiguang , Geng Guoqiang & Shao Hu . ( 2025 ). A Study on Coal Price Forecasting Integrating XGBoost and Graph Convolutional Networks . Coal Economics Research , 45 ( 02 ), 39 – 46 .