透過您的圖書館登入
IP:18.226.222.12
  • 學位論文

台電顧客族群用電量特性分析研究

The Analysis of Power Usage Characteristics of Taiwan Power Company’s Customer Groups

指導教授 : 黃怡詔

摘要


台灣電力公司在台灣是主要的電力提供者,但隨著經濟的發展,台灣在尖峰時間的用電負載量也隨之增高,為了解決此問題,台電公司制定了時間電價策略,以抑低在尖峰時段之電力負載量;近年來隨著電力市場逐漸的開放,未來產生的競爭也愈來愈大,台電公司逐漸意識顧客關係管理的重要性,各區營業處設置專員進行採訪客戶、解答客戶問題的工作;然而,唯有將客戶用電屬性加以分群,並且制定出適合各分群的合理電價,才能完成顧客關係管理並有效率的服務顧客;本研究使用群集分析方法中的二階段分群法,其結合了華德法及k-means分群法,以台電公司用戶服務資料倉儲系統的資料,對台電公司客戶的用戶特性進行分群,並且將其用電特性加以分析, 提供給台電公司作為日後服務的參考資料。

並列摘要


The Taiwan Power Company (T.P.C) is the major power supplier in Taiwan, and as the development of the economics, the peak time’s power load is thus accelerate. For solving this issue, the T.P.C starts to formulate the plan of time-of-use electricity rates to restrain the power load during the peak time. In recent years, with the gradual opening up of the electricity market, competition in the future will become more and more intense. Therefore, the T.P.C is gradually realizing the importance of the customer relationship management. Consequently, each branch of the company arranges the commissioners to visit the customers, and to give answers to solve the customer’ problems. Nevertheless, there is only distinguishing the attributes of the customer, hence the customer relationship management can be accomplished as well the customers can be served efficiently, and then to formulate a reasonable electricity rate. In this research, the two-stage clustering method of cluster analysis technique is use to combine with the Wards Method and k-means clustering method. Furthermore, we made use of the Data Warehousing System of the User Service of T.P.C to cluster user’ characteristics; and to analyze their power usage Characteristics. In conclusion, all of these data can be provided to the T.P.C as the reference for the future service.

參考文獻


5.林誠、劉福堂,「資料探勘在寬頻網路客戶目標行銷之應用研究」,電子商務學報,第七卷,第二期,第121-138頁(2005)
1.王建智、陳銘樹、徐正容,「應用資料探勘中的分群分析技術來探究高階健檢客戶之型態組成以進行客戶關係管理-以台北某醫學中心之高階健檢客戶為例」,健康管理學刊,第六卷,第一期,第49-59頁(2008)
4.周歆凱、張怡秋、黃興進、蔡明足、翁林仲、蘇喜,「運用購物籃分析技術探討滯留急診超過24小時病患特性」,醫務管理期刊,第八卷,第三期,第216-231頁(2007)
40.Yan-Ping Huang, Chung-Chian Hsu, “Concept hierarchy-based pattern discovery in time series database: a case study on financial database,” Journal of Information Management, 14(s), 219-237 (2007)
8.張心馨、蔡獻富,「以Data Mining技術結合SOM和K-Mean的消費者分群方法於顧客關係管理和績效獲利性評估之實證研究」,資訊管理學報,第十一卷,第四期,第161-203頁(2004)

延伸閱讀