隨著能源議題受到關注,除了推廣永續能源之外也需要探討節約能源的方針。一般家庭生活用電屬於最普遍的電力消耗,藉由分析用電資料查找用電過高亦或是浪費電力之行為,透過住戶知曉其用電行為後自發性省電以求節約能源目的,但是,住戶無法從單純的總用電量得知數據涵蓋資訊,因此本研究提出一套資料倉儲系統以求後續對用電資料進行分析管理並提出適合之決策。 處理用電資料前,擁有格式清楚且完整的資料集合尤其重要,本研究將所收集已公開的用電資料集合提出欄位設計不當、量測項目不清楚及紀錄不完整等問題,若無統一格式將造成資料於後續分析上的困難,因此本研究將用電資料集合之屬性資訊相互比較並提出適合之格式,以作為用電資料的共同語言。 本研究使用REFIT英國公開用電資料集合提出一套資料倉儲系統,藉由維度屬性資訊與用電資料之關聯性設計各維度表結構,其中包含時間、環境、建築資訊、地址及電器共5個維度表。本研究提出此資料倉儲系統目的為讓使用者對用電資料進行彈性的維度分析,並讓使用者知悉用電數據之重要資訊,由於資料倉儲為結構化資料的特性,此系統可供使用者查詢用電資料的趨勢,透過趨勢情況得知住宅用電行為,再近一步規劃後續決策,甚至讓住戶以自發性規劃其節電行為,以達節約能源的效益。 ;With the importance of energy issues, how to make the occupants of a residential or commercial building more aware of their own electricity consumption and then change their behavior to save electricity has always been the top priority. Today, countries are gradually promoting the installation of smart meters. The literature also has a collection of publicly available power consumption data sets for selected countries. With reference to the national electricity data format, it is necessary to clarify and design the most suitable method for Taiwan, so as to meet the needs of various analysis in the future. Furthermore, this study uses the UK REFIT public data set, transforms the data format requirements for data warehousing analysis, uses Microsoft SQL Server’s data warehousing tools, designs the star schema as well as dimensions and fact tables, and finally presents data warehousing results in the form of trend analysis. In addition to understanding the people′s electricity habits and designing energy-saving behaviors, the identified patterns can be extended and applied in the smart meter data sets from Taiwan, so as to plan dynamic energy-saving behaviors to achieve energy conservation.