考慮室內空調舒適度之需量反應與訂定最低抑低契約容量

Demand Response with Considering Indoor Thermal Comfort of HVAC and Determination of the Minimum Suppression Contract Capacity

劉禹均(Yu-Chung Liu)

指導教授 ： 張永宗

國立臺北科技大學/機電學院/能源與冷凍空調工程系碩士班/碩士(2012年)

摘要

建築物內的空調負載隨時在變動，若要執行需量卸載，又要兼顧室內舒適度，而只對冰水泵、空調箱、風機等空調周邊設備做頻率控制進行卸載，則要決定一個適合與台電公司簽訂的抑低契約容量相當困難。若能事先得知建築物內的空調負載，即可估算出空調設備所卸載的量。
本研究以兩個案例來驗證利用類神經網路預測建築物內24小時之空調負載的準確度，而空調負載之預測值與實際值的平均誤差率僅3.56%。本研究也進行了兩次卸載實驗，在尖峰用電期間將室內舒適度維持在PMV = 0~0.5及PPD = 5~10的理想範圍內，並執行需量卸載控制。以預測實驗當天24小時之空調負載與需量卸載後實際空調負載做分析及比較，進而估算出適合與台電公司簽訂的「最低抑低契約容量」。以此方法對中央空調系統進行需量卸載控制，除可維持室內舒適環境亦可達到減少空調系統的電費支出及節能減碳之目的。

關鍵字

需量卸載 ； 類神經網路 ； 負載預測 ； 舒適度 ； 建築能源管理系統

並列摘要

The cooling load of a building will be changed anytime. To enforce demand load-shedding and to give consideration to indoor thermal comfort, it is difficult to make a decision of a fitting suppression contract capacity that sign with Taiwan Power Company through controlling the frequency of air conditioning facilities such as chilled water pump, air handling unit and fan coil. If the amount of cooling load of a building could be learned in advance, then the load-shedding of air conditioning facilities could be estimated.
This study used two cases to test and verify the accuracy of forecasting the cooling load of a building in 24 hours by Artificial Neural Network. The mean error between forecasted and actual cooling load is only 3.56%. This study also proceeded load-shedding experiment twice. The author maintained the indoor thermal comfort to an ideal range that PMV was from 0 to 0.5 and PPD was from 5 to 10 during peak time and enforced demand load-shedding control. The author analyzed the data and compared the forecasted 24-hour cooling load of the experiment day with the actual cooling load after demand load-shedding. This enable the author to estimate the fitting “minimum suppression contract capacity” that sign with Taiwan Power Company. The application of this method in order to enforce demand load-shedding control with the central air conditioning system can not only maintain the indoor thermal comfort but also achieve the goal of reducing electricity costs of air conditioning system, energy conservation and carbon reduction.

並列關鍵字

Demand unloading ； Artificial neural network ； Load forecasting ； Thermal comfort ； Building energy management system

參考文獻

[6]盧浩仁，需量反應中卸載最佳化之設計，碩士論文，國立臺北科技大學自動化科技研究所，台北，2005。

[4]陳志豪，應用類神經網路預測空調負載，碩士論文，國立臺北科技大學能源與冷凍空調工程研究所，台北，2004。

[5]Bing Donga, Cheng Caob, Siew Eang Leea, “Applying support vector machines to predict building energy consumption in tropical region,” Energy and Buildings, 37 (2005) 545–553.

[7]林良興，應用適應性類神經模糊推論系統於冰水主機耗電量之預測，碩士論文，國立臺北科技大學能源與冷凍空調工程研究所，台北，2008。

[8]Qiong Li, Qinglin Meng, Jiejin Cai, Hiroshi Yoshino, Akashi Mochida, “Predicting hourly cooling load in the building: A comparison of support vector machine and different artificial neural networks,” Energy Conversion