目前台灣電廠仍以高汙染的火力發電及高危險的核能為主，屬於集中式發電。近年來，政府除了積極發展再生能源外，也努力將電網轉型朝向分散式發電。為了解決缺電的問題及避免擴增電廠的考量下，政府開始反思由電力需求端進行能源管理。為了使電網的備轉容量更加穩定，台電提供不同的「時間電價」的方案，引導用戶改變用電行為，將用電從高峰轉移至非高峰時段，未來也持續計畫提供更多的電價方案推行至住宅區，以達到削峰填谷之目的。因應能源轉型所帶來的衝擊，推動智慧電網發展儼然成為當務之急。然而，智慧電表在電網中扮演關鍵的角色，並預期結合時間電價調整民眾用電行為，使其發揮最大之效益。因此，在可見的未來裡，家庭能源管理將越顯重要。 基於上述未來趨勢及政策考量，本研究將問題定義在台電所提供的時間電價方案下，進行家用電器排程規劃，並且建立數學模型，使用混合整數規畫求解。本研究的目的為找出合適的家用電器操作時間，同時將使用者的用電行為考慮進模型中。在研究的情境設計上，此研究經由多次的家庭訪談後，採用台灣常見的家用電器與電器使用方式作為參考資料。再利用電腦實驗的方式得到結果，此研究將會分析不同種類的電器對於整體電價的影響程度，並且針對使用者可接收不方便性的容忍度進行調整分析。 有別於其它的相關研究，此研究的主要貢獻在於量測使用者不方便性的想法創新。依據不同使用者的使用喜好，利用不方便性矩陣與不方便性可接收容忍度的設計，使數學模型更貼近真實使用的情況。另外，本研究將會根據實驗結果，對於採用時間電價的使用者給予實質上的使用建議，並且在本研究的最後討論未來相關可行的延伸研究。 ;At present, Taiwan’s main energy supply consists of thermal and nuclear power generated in a centralized manner. The government has established plans to decentralize power generation and incorporate renewable energy into the power system. To resolve power shortage while avoiding building more power plants at the same time, the government has begun to manage energy consumption from the demand side. By using the Time-of-Use (TOU) pricing to alter consumer behavior, the government can better stabilize operating reserves during energy shortage. In addition, Taipower has plans to offer TOU pricing scheme in the residential sector. In smart grid (SG), smart meters play an important role in enabling TOU. Once the use of smart meters has expanded, home energy management will be more important. Consequently, this research studies the problem of scheduling household appliances under TOU, and builds a mathematical model using mixed-integer programming to solve the problem. The purpose is to find appropriate scheduling under TOU pricing, while taking consumer behaviors in household appliances into consideration. This research concerns a typical Taiwanese family’s electricity consumption behavior as the basis for its scenario. The computer experiment for this research analyzes the relationship between the different types of appliances and electricity cost fluctuations, and adjusts the acceptable inconvenience range. The contribution that this research makes is for a new concept of measuring inconvenience. By using an inconvenience matrix and acceptable tolerance level, decided by consumer preference, it makes the model more realistic and humanized. At the end of the paper, this research will give the consumers substantial suggestions to choose TOU pricing, and discuss the extent of the research in the foreseeable future.
