:::

臺灣能源期刊論文全文

臺灣能源期刊第7卷第4期內容

出刊日期:December, 2020

題目
需量反應基準線設計
Title
Demand Response Baseline Design
作者
蔡昊廷、吳國賓、陳俊宇、梁佩芳
Authors
Hao-Ting Tsai, Gwo-Bin Wu, Chun-Yu Chen, Pei-Fang Liang
摘要
基準線的設計在需量反應中扮演很重要的角色,一個不好的基準線設計,不僅會影響用戶參與的意願,也很容易受到有心人士的操弄並賺取大量利益,且對於系統的幫助有限。因此,本研究針對國內需量反應中的負載管理來討論其效益計算方式,由北美能源標準委員會(North American Energy Standards Board, NAESB)的定義可以了解大部分電力調度中心(Independent System Operator, ISO)是利用第一類型基準線(Baseline Type I - Interval Metering)來對該類型服務進行效益計算,因此收集北美各ISO在該方法下的做法,並利用國內用戶用電資料對於國外第一類型基準線的做法進行基準線計算測試。澳洲能源市場調度中心(Australian Energy Market Operator, AEMO)提出了三種檢定方法,用來檢驗基準線的準確性、偏差性以及變異性,分別是相對均方根誤差(Relative Root Mean Squared Error, RRMSE)、平均相對誤差(Average Relative Error, ARE)以及相對誤差比率(Relative Error Ratio, RER),利用上述檢定方法來判斷用戶較合適的基準線計算方式。以本研究之結果來看,該用戶在當日負載調整為具有20%限制的比例法下的RRMSE與RER皆最低,代表在該方法下的基準線較為準確。
關鍵字
需量反應,基準線設計,相對均方根誤差
Abatract
The baseline design plays an important role in the demand response. A bad baseline design is not only affects the willingness of users to participate, but also easily manipulated by people with bad intention and earns a lot of benefits, and it has limited help to the system. The study discusses the performance calculation method for the load management in the demand response. According to the definition of the North American Energy Standards Board (NAESB), most ISOs use the Baseline Type I - Interval Metering to calculate the performance. Therefore, collect the calculation methods of the Baseline Type I - Interval Metering in North American ISOs, and use the AMI data to calculate user’s baseline by different methods. The Australian Energy Market Operator (AEMO) has proposed three verification methods to test the accuracy, deviation, and variability of the baseline, called Relative Root Mean Squared Error (RRMSE), Average Relative Error (ARE), and Relative Error Ratio (RER). This study uses these verification methods to find a most appropriate baseline calculation method. Based on the results of this study, the RRMSE and RER under the proportional method with a 20% limit baseline adjustment are the lowest, which means that the baseline under this method is more accurate.
Keywords
Demand Response, Baseline Design, Relative Root Mean Squared Error.