主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

Chinese Journal of Management Science ›› 2015, Vol. 23 ›› Issue (8): 132-138.doi: 10.16381/j.cnki.issn1003-207x.2015.08.015

• Articles • Previous Articles     Next Articles

Research on the Consumption Fluctuation Characteristics and Customer Segmentation of Natural Gas Customer

HE Zhi-yi1,2, CHEN Zheng-hui1,3   

  1. 1. Antai College of Economics & Management, Shanghai Jiao Tong University, Shanghai 200052, China;
    2. Newhuadu Business School, Fuzhou 350108, China;
    3. PetroChina West East Gas Pipeline Company, Shanghai 200122, China
  • Received:2014-02-26 Revised:2014-07-30 Online:2015-08-20 Published:2015-08-19

Abstract: It is a realistic problem for the production suppliers that how to effectively meet the demand of customers who have continuous consumption and the consumption are cyclically fluctuated. Based on the customers' daily natural gas consumption time series data, the spectral methods are applied to analyze the fluctuation characteristics of consumption. Then, the discriminant function equation is constructed using hierarchical clustering analysis and discriminant analysis according to the volatility indicators. Based on the Jiangsu Province customers' daily consumption data of "WE" natural gas sales company, this paper found that: The fluctuation characteristics can be depicted by 13 indicators derived from spectral methods analysis. The customers can be classified by those fluctuation indicators and adopt different pitch-peaking method for different categories of customers. An analysis foundation for natural gas suppliers optimizing their pitch-peaking project is provided. In addition, this method can be used on new customers' segmentation according to their future natural gas demand fluctuation so that the suppliers can proactively take effective measures to solve the problem of consumption fluctuations of new customers.

Key words: natural gas, spectral methods, cluster analysis, discriminant analysis, customer segmenting

CLC Number: