基于数据挖掘技术的天然气价格预测方法研究

天然气价格是影响天然气企业经营决策与运营效益的重要因素,在此背景下,如何准确地预测未来天然气价格自然成为产业界关注的热点话题。此外,在数据挖掘技术快速发展的时代,如何将该技术应用于传统的天然气行业,融入天然气价格的预测当中,也是学术界所探讨的重要话题。基于此,本文首先回顾了以往天然气价格预测方法,然后以传统数据挖掘技术中的模式序列相似性搜索方法(PSS)为基础,通过对该方法中历史序列搜索匹配机制及结果处理机制的改进,提出了一种新的改进模式序列相似性搜索(APSS)天然气价格预测方法。在方法构建之后,采用美国天然气日度现货价格数据对该方法的有效性进行了实验验证。实验结果表明,本文提出的基于数据挖掘技术的APSS方法能够实现对天然气价格的合理预测,且与传统的PSS方法相比,APSS方法的预测结果具有更高的预测精度。

Research on the Forecasting Method for Natural Gas Price Based on the Data Mining Technique

Natural gas price is a key factor which influences the operation decision and profit of natural gas enterprises. In this case, how to forecast the price accurately has naturally become a hot topic in the natural gas industry. Besides, with the rapid development of data mining techniques, how to use the data mining techniques in natural gas price forecast also attracts lots of attentions from the academia. Firstly, the existing forecast methods of natural gas price is reviewed in this paper. Secondly, by modifying the two mechanisms in traditional pattern sequence similarity search (PSS) which is based on data mining techniques, i.e., the matching mechanism in searching historical data series and the result processing mechanism, a new adjusted pattern sequence similarity search (APSS) method is proposed in this paper to forecast the natural gas price. Thirdly, to verify the validity of the proposed method, the data of US daily natural gas spot price are used to the method. The empirical results show that the proposed APSS method can forecast the natural gas price reasonably and the forecast results of the APSS method have a higher prediction accuracy by comparing with the traditional PSS method.