應用模擬退火法於空調系統運轉之最佳化

Simulated Annealing for Optimal Operation of Air-Conditioning System

李俊賢(Jun-Xian Li)

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(2009年)

https://doi.org/10.6841/NTUT.2009.00491

摘要

現代的科技工業競爭相當激烈，若能減少能源支出費用，其產業的競爭力勢必能提高，而當中空調耗電量所佔比例最大，從這領域進行節能，應該是最快收到成效。在空調系統的總耗電量中，冰水主機耗電約佔60%，水泵系統約佔20%，空氣側設備也約佔20%，由於冰水主機耗電佔了全系統一半以上，因此目前空調系統最佳化的研究通常集中於冰水主機控制上，較少著眼於其他部分，即使於各元件上加裝變頻器，也多以經驗去調整其設定值，並無法達到其最佳節能功效。
本研究引用外國學者所推導出的熱平衡方程式，再以迴歸分析建立各元件的耗電量模型，針對空調系統中冰水主機、空調箱、水泵相互影響的關係，使用模擬退火法(SA)來尋找出冰水出水溫度、風量和水量的最佳設定點，SA是種結構簡單，隨機性的演算法，無論目標函數是否為連續或可否微分，皆可適用，且會利用機率函數和控制參數「溫度」來判斷是否接受該解為新解，因此提供SA有能力去跳脫區域最小值，本論文以系統總耗電量為目標函數，在滿足空間負載的冷凍能力條件之下，完成整體空調系統的最小耗電量。而由結果中也可確實觀察到，尋找此三種變數之最佳設定點，其耗電情形與未做改善前相比較，耗電量可降低17.15%，顯示本研究確實可找出空調系統運轉之最佳設定點，達到節能之目的。

關鍵字

最佳化運轉設定 ； 模擬退火法 ； 迴歸分析 ； 空調系統最佳化 ； 熱平衡方程式

並列摘要

The technology industry is facing fierce market competition today. If companies in this industry are able to cut back on energy expenses, their competitiveness will be substantially uplifted. Among the energy expenses, power consumption from air-conditioning takes up the largest portion; therefore, cutting back expenses by lowering the power consumption from air-conditioning should yield instant effect. In an air-conditioning System, the chiller consumes 60% of energy, the pump system 20%, and the air side facilities 20%. Since the chiller takes up over half of the power consumption, the current researches on air-conditioning system energy optimization concentrate mainly on the control of the chiller. The rest parts of a system are hardly researched because energy optimization can hardly be achieved as the configuration is often adjusted through experience even when an inverter is installed.
This research adopts the heat transfer equation developed by foreign scholars and uses regression analysis to establish the power consumption model of each component. Simulated Annealing algorithm (SA) is then used to find the best configurations of chilled water supply temperature, supply air flow, and water flow, targeting on the interrelationships among the chiller, air handing unit, and zone pump. SA is an algorithm with simple structure and randomness. It is applicable to a wide range of calculations, no matter whether it is a serial objective function or it can be differentiated, and probability function and control parameter-“temperature” can be used to determine if the results are accepted as new results. Therefore, by supplying the SA method, the algorithm is enabled to break away from the local minimum. This thesis configures the total power consumption of the system as the objective function and achieves the minimum overall air-conditioning system power consumption under the criteria of achieving the cooling capacity that satisfies the loading requirement. From the results, we find that power consumption is lowered by 17.15% if we compare the power consumption of the system using the three best operation parameters derived in this research to a system without any adjustments. It shows that this research indeed uncovers the best parameters for air-conditioning system operation that achieves the objective of power-saving.

並列關鍵字

best parameters for air-conditioning system operation ； Simulated Annealing ； regression analysis ； air-conditioning system energy optimization ； heat transfer equation

參考文獻

[3] 魏仁杰，應用基因規劃法於冰水主機負載分配之最佳化，國立台北科技大學能源與冷凍空調工程系碩士論文，民96。

[7] 傅定儀、趙宏耀，電子廠節約能源方向與案例說明，中國冷凍空調雜誌，1996年2月，第79-89頁。

[10] 戴邦文，應用於節能績效保證合約之冰水主機性能驗證方法，國立台北科技大學能源與冷凍空調工程系碩士論文，民95。

[11] 江建樺，應用模擬退火演算法於冰水主機負載分配之最佳化，國立台北科技大學能源與冷凍空調系碩士論文，民94。

[16] 洪顥庭，高轉速離心鼓風機設計參數對性能影響，國立台灣大學機械工程所碩士論文，民97。