Natural gas consumption behavior of companies by clustering analysis
Yükleniyor...
Tarih
2021
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Pergamon-Elsevier Science Ltd
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
We have still consumed natural gas as a restricted source of energy in our daily life. Moreover, the consumption of natural gas energy will continue to increase by the year. Although many studies have focused on electrical energy consumption, natural gas is another significant energy source that can be examined. Since companies consume much more gas, their gas consumption data are examined in this study. The study contributes to the literature by applying intuitionistic fuzzy c-means (IFCM) methodology to the natural gas industry. The main motivation and advantage of the methodology is two-fold. Because of its fuzziness, one data point can be assigned into more than one cluster, similar to real-world cases. Because of its intuitionistic side, it considers membership, non-membership and hesitant degrees. These two strengths of IFCM improves the clustering accuracy. IFCM clustering was used to arrange the companies with respect to the consumption amount to increase the understandability because 1049 companies' consumption data were collected. A calendar view was developed to visualize the consumption amounts in the clusters. The changes in consumption amounts were presented in different weather temperatures. Whereas some clustered companies were directly affected by temperature changes, others were not affected. The companies in the clusters were analyzed with respect to two main criteria: regularity and complexity. The findings showed while high levels in routine are related to manufacturing companies, high complexity level is an indicator of being active in the service industry.
Açıklama
Anahtar Kelimeler
Company behaviors, Natural gas consumption, Intuitionistic fuzzy c-means, Calendar view, SCADA smart metering, Energy-Consumption, Electricity Consumption, Load Profiles, Prediction, Algorithm, Classification, Implementation, Intelligence
