用戶負載特性為電力系統最基本之資料，透過對用戶端負載特性之了解，將能有效提高負載預測之可靠性，改善電源開發系統規劃及電力系統經濟調度。此外用戶之負載特性，將可規劃符合成本效益之節約電能及負載管理方案，以降低電力系統容量不足之壓力。
本文以負載特性調查研究為基礎，首先以分層隨機抽樣法決定抽樣用戶數，於抽樣戶裝置智慧型電錶，量測記錄用戶全年每15分之耗電量。將所收集之電錶資料作異常資料處理後，以統計分析方法推導出各行業別用戶正規化日負載模型。同時配合台電用戶資訊系統(CIS)之行業別用戶之售電量，推估北市及高雄兩區處之行業別負載組成。並應用統計迴歸分析方法擬合各行業別用戶用電量之溫度迴歸模型，探討其用電量受溫度變化影響之程度，以評估各行業別用戶及整個台電系統由於溫度變化所造成之負載變動量。
負載特性調查對電力公司之重要性將與日俱增，負載資料也將大量被應用於電力事業的各相關部門，作為改善營運績效之重要參考數據。本論文探討用戶負載之溫度敏感度分析，可探討未來台電推動空調負載管理以降低系統尖峰負載之潛力分析。

Customer load characteristics plays the ndamental role for more reliable load
forecasting. It can also be used to enhance the system expansion planning and economic dispatch more effectively. Besides, the system capacity shortage due to peak loading can be relieved by the strategy of energy conservation and load management with customer load models.
A systematic procedure is proposed in this thesis to study the effect of temperature change to the power system load demand by using the typical load patterns of customer classes. The billing data of all service customers are retrieved to derive the daily load profile of the selected Taipower district. To verify the accuracy of the estimated load composition, the simulation results are compared to the actual load profile collected by the SCADA system. The sensitivity analysis of load demand with respect to the temperature change for each customer class is performed by statistic regression according to the actual customer power consumption and temperature data. With temperature rise, the load contribution by each customer class is updated by the corresponding temperature sensitivity and integrated together to form the new load profile of the service district.
In the future, the load research will play more important role for power utility companies. Load data will be utilized to a greater extent by various departments in utility companies. For instance, the proposed load survey system can solve the customer load characteristics more accurately to support various applications. By refer the temperature sensitivity analysis based on the customer load research, can evaluate the potential of air conditioner load management to reduce the system peak loading can be inhibit. With this information, the proper incentive of cycling control of air conditioners can be designed to achieve more effective load management.

摘要
目錄
圖表索引
第一章 概論
第二章 負載特性調查之用戶抽樣
第三章 數位電錶系統架構及用戶異常用電資料處理
第四章 負載組成推估模式及應用
第五章 行業別負載之溫度敏感度分析
第六章 台電系統負載之溫度敏感度分析
第七章 結論與未來研究方向
參考文獻

參考文獻
[1]"Load Research Manual”, Association of Edison Illuminating
Companiesm,1990,2.
[2]"台電系統中部級北部地區負載特性調查方析研究”,台電公司八十七年度研究計劃,台灣
電力公司,1998/5。
[3]C.F. Walker and J.J. Pokoski, “Residential Load Shape Modeling Based on
Customer Behavior” , IEEE Trans. on Power Systems, Vol. PAS-104,No.7 , July
1985, pp.1703-1711.
[4]C.S.Chen, J.C.Hwang and C.W.Huang, “Application of Load Survey Systems to
Proper Tariff Design”, IEEE Trans. on Power Systems, Vol. 12, No. 4, Nov.
1997, pp.1746-1751.
[5]C.S.Chen, J.C.Hwang and C.W.Huang, “Determination of Customer Load
Characteristics by Load Survey System at Taipower”, IEEE Trans. on Power
Delivery Vol.11 No.3, July 1996, pp1430-1435.
[6]C.S.Chen, J.T.Leu, “Interruptible Load Control for Taiwan Power Company”,
IEEE Trans. on Power Systems, Vol.5, No.2, May, 1990, pp460-465.
[7]台電公司行業別分類表.
[8]趙坤芳, “SAS基本資料處理與操作”,全華,1998，7
[9]台電系統負載特性調查分析研究,第三期計劃,第二次期中報告 第一冊。
[10]儲全茲,“抽樣方法”,三民書局,1993,3.
[11]陳順宇,“多變量分析”,華泰,1998,7.
[12]謝暉光,“抽樣調查之本質與實務”,科學月刊,1981.
[13]唐明月,陳英亮,“機率學與應用”,中興管理顧問公司,1984,9.
[14]魏應澤,“抽樣原理及其應用”,三民書局,1977
[15]顏月珠,“商用統計學”,三民書局,1990,6.
[16]林惠玲,陳正倉,“統計學方法與應用”,上冊,雙葉,1999,11.
[17]Bailey and Daniel, “Probability and Statistics Model for Research”, John
Wiley, 1971.
[18]章立民,“Visual FoxPro6.X中文版程式設計”,松崗,1999,4
[19]吳宗正,“迴歸分析—理論與應用”,復文書局,1995,3
[20]孫育義,“或然率與統計”,興業圖書股份公司,1975,5.
[21]張紹勳,“統計分析—高等統計”,松崗,1995,8.
[22]黃文隆,“抽樣方法”,滄海書局,1998,12.
[23]黃鐘慶,“負載特性調查在合理化電價訂價策略之應用”,國立中山大學電機工程研究所
博士論文,1995/5.