頁籤選單縮合
題 名 | 電力耗能負載預測與節能應用=The Energy Saving Application of Electricity Load Forecasting |
---|---|
作 者 | 陳束弘; 林政廷; 蔡宗成; 張語軒; | 書刊名 | 臺灣能源期刊 |
卷 期 | 1:5 2014.12[民103.12] |
頁 次 | 頁601-612 |
分類號 | 448.115 |
關鍵詞 | 電力負載預測; 耗能分析; 節能應用; Electricity load forecasting; Energy consumption analysis; Energy saving Applications; |
語 文 | 中文(Chinese) |
中文摘要 | 本論文透過電力負載預測,提供未來24小時至一星期每小時的耗電量及用電趨勢,使用者可以提前預知未來的用電負載狀況,來達到耗能提醒的功能,以促使用電戶產生自發性節能意識,依據預測結果進行用電設備監控來達到節能的目的。 本論文使用支持向量機演算法進行用電負載預測。比較真實用電資料及預測結果,可以觀察出二者的用電趨勢是一致的,其平均絕對值誤差率(MAPE)在5%以下,是屬於高準確的預測。因此藉由預測結果的提供,使用電戶了解其用電習慣,當用電戶發現未來可能會有高耗能發生時,可提早進行節能策略的規劃與進行,進而達到節能的目的。 |
英文摘要 | In this paper, forecasting for 24 hours-ahead to 7 days-ahead of hourly electricity load and demand trend is developed, and the projection outcomes will be offered to users as notifications to raise their will of voluntarily energy saving. In addition, appliance monitoring and control will be applied in accordance with the projection results to achieve the goal of energy saving. Support vector machine (SVM) algorithm is chosen to accomplish the short-term electricity load forecasting, and the outcome of electricity forecasting fits the trend of real power demand. In addition, the forecasting is highly accurate, the average error rate is below 6%. By offering the forecasting results to the end users, patterns of electricity consumption can be discovered. When the users is notified that heavy energy demands are going to be occurred, they can make energy saving plans in advance to achieve power demand reduction. |
本系統中英文摘要資訊取自各篇刊載內容。