請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/8266
標題: | 結合使用者自我節能意識於多人智能家庭的需量電力管理系統 Demand Side Management System with Self-Awareness of Energy Saving for Multi-User Smart Home |
作者: | Jheng-Huang Hong 洪正皇 |
指導教授: | 傅立成(Li-Chen Fu) |
關鍵字: | 需量管理系統,家庭區域網路,強化學習,深度Q-網絡,電器可控性,用戶行為,節能動機,智慧電網, Demand side management,home area network,reinforcement learning,Deep Q-Network,appliance flexibility,user behavior,self-awareness of energy saving,smart grid, |
出版年 : | 2021 |
學位: | 碩士 |
摘要: | 隨著全球能源消耗上升,如何有效的使用能源變成一個重要的議題,需求端管理系統是一項能用來解決此議題的技術。對於每一間建築,一天中會有用電的高峰時段,這些用電高峰會導致電源供應方的負擔,在時間電價機制的鼓勵下,需求端管理系統可以協助使用者將非立即需要的用電轉移至離峰時段。另外,隨著物聯網技術的發展,家中可連網的裝置能夠形成一個家庭區域網路,使得管理系統能夠監控家中的整個狀況,並協助使用者控制電器節省能源。 在這篇論文中,我們著重在住宅型房屋,實現了能夠應用在多使用者智能家居中的需量管理系統,其中包含了個人化、自動化以及節能動機三大概念,首先識別正在使用裝置的使用者身分,所有服務都將根據使用者個人習慣運行。在自動化的部分,透過多個深度強化學習代理們,分別管理不同的電器、能源儲存系統及再生能源,識別使用者的當前行為以更準確的控制所有電器。除此之外,我們使用最大似然估計來估計電器的使用時間分佈,並計算各項電器的彈性度以量化使用者使用電器的規律程度。為了進一步節省能源,節能模組會提供使用者節能建議,節能建議的時機及內容會根據個人節能分數和電器彈性度決定,並且考慮到多使用者家庭的情況。同時,個人化的能源使用情況也會經由視覺化界面提供給使用者,這些能源反饋資訊能夠提高使用者的自我節能意識。 在實驗結果中,當提供能源反饋後,使用者會願意主動執行更多節能行為,進而降低更多能源花費。此外,根據使用者體驗調查結果,所提出的系統在實用性及易用性上都取得使用者的認同。 As global energy consumption keeps increasing, the energy efficiency has become an important issue. Demand side management (DSM) system is a technology that can be used to solve this issue. In many houses, there are some time periods in a day when consumers have high electricity demands, and these peak demands will cause a burden on the smart grid. In order to solve energy efficiency problem, the demand side management system can help users to postpone non-urgent demands to off-peak hour subject to the mechanism of real-time pricing policy. In addition, due to great advancement of Internet of Things (IoT) technology, the devices which have internet module can all be connected into a home area network, enabling DSM to monitor the entire state of house and to control appliances. In this thesis, we focus on a residential house and implement a demand side management system that can be applied to multi-user smart home, which includes three important elements: personalization, automation, and energy saving motivation. First, the system will recognize the identity of all users, which facilitates personalized services to be provided in a multi-user smart home. In automation part, multiple reinforcement learning agents are developed to separately manage different appliances from different categories, energy storage system, and renewable energy resource so that all devices are under accurate control. Moreover, we use maximum likelihood estimation to estimate the usage time distribution of various appliances, and also evaluate their flexibility in order to quantify the regularity of appliance usage of every user. According to appliance flexibility, the energy saving module will suggest users to undertake more energy saving actions. The timing and content of these suggestions will be determined by our proposed energy saving suggestion decision (ESSD) algorithm. Meanwhile, the information of the personalized energy usage will also be provided to users through a visual interface. Such energy feedback information can improve the users’ self-awareness of energy saving. In the experimental results, DSM that integrates automation and energy feedback helps users decrease energy costs more than DSM with only automatic smart control. Throughout the experiment, we notice that energy feedback dose let users be more aware of need for energy saving and hence reduce more energy saving behaviors. Lastly, according to user experience survey, users agree that the design of the proposed system is practical and easy to use, with an average satisfaction score being 4.4 out of 5. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/8266 |
DOI: | 10.6342/NTU202100623 |
全文授權: | 同意授權(全球公開) |
電子全文公開日期: | 2024-03-01 |
顯示於系所單位: | 資訊工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
U0001-0502202123574700.pdf | 3.75 MB | Adobe PDF | 檢視/開啟 |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。