要讓人們養成省電的習慣,除了利用科技去量測日常生活所消耗的電量,更需要 將這些量測資料聯結到人們的生活作息。在這篇碩士論文提出一套居家活動用電 分配量測系統,幫助計算每一項居家活動(例如煮飯或看電視)所消耗的電量。 論文中將詳述電量系統的設計理念及實作細節,並將系統真實的佈置在 3 個家庭、 共 9 位授試員、為期 4 個星期的實驗以進行系統效能評估。實驗中佈置了 57 個 測量點,共收集 1200 多個有用電的居家活動。此用電分配量測系統能正確的辨 認出 92.97%的居家活動,並且每個活動的用電量測量準確率高達 95.55%。
Connecting power-monitoring data to everyday human activities is an important aspect of promoting energy-saving behavior. This thesis proposes an activity level power monitoring system (ALPS) that disaggregates energy consumption based on human activities (e.g., cooking and watching TV). This thesis reports the design and implementa- tion of the ALPS, and evaluated the proposed system by deploying 57 power monitoring nodes in three homes with nine participants over a 4 week period and collected over 1200 household energy-consuming activities. Results show that the proposed system successfully recognizes 92.97% of energy consuming activities, and achieves an accuracy of 95.55% in activity level energy monitoring.