由於能源逐漸耗竭潔淨再生能源的技術未獲突破,提升能源使用效率是最具效益的節決方案,而蒸發式冷卻(Evaporative Cooling)應用於空調系統能降低冷凝溫度提升系統性能,達到提升能源使用效率之目的;而現今R22冷媒因為具有破壞臭氧層潛力在不久將停用,近年來歐盟、美國及日本紛紛改採環保冷媒R410A取代。 本研究主要建立冷凍循環中各元件之數值模式,以Excel內建之VBA撰寫成蒸發式冷卻空調系統模擬程式,最後以倒傳遞類神經網路(BPN , Back-Propagation Network)作為系統元件之設計,本研究並有組裝5RT之蒸發式冷卻空調系統實驗機R22及R410A各一台,以實際運轉性能與軟體模擬結果互相比對驗證,再以本程式模擬數據訓練倒傳遞類神經網路。 由經驗公式及倒傳遞類神經網路模擬結果比較僅有0.6%RMS誤差,由此可發現倒傳遞類神經網路適用於蒸發式冷卻空調系統模擬,從模擬結果得知R410A不僅無破壞臭氧層潛力,且在相同外型尺寸下性能優於R22 9.8%,在相同製冷能力下熱交換器可縮小25.3~43.7%。
Because of energy shortage and renewable energy technologies have not been breakthroughs, to enhance the energy efficiency is the most effective solution. Applying evaporative cooling technology can reduce the condensing temperature and increase system performance, and then reach the purposes of energy efficiency. On the other hand refrigerant R22 will be suspended in the near future because of potential damage to the ozone layer. Recently European Union, the United States and Japan have adopted refrigerants R410A to replace R22. This research establishes the numerical model of refrigeration cycle components from thermodynamics and heat transfer, and develops the evaporative cooling air-conditioning system simulation program in Excel VBA as well as uses back-propagation neural network system components for the system optimization design. This research also assembled a 5 RT evaporative cooling of air-conditioning system using R410A and R22. By the experimental investigations the simulation results were compared and the present model was verified, and then using the simulation data to train the BPN model. The results obtained from empirical formula and back-propagation neural network simulation were compared and the errors were with in 0.6%. It shows that back-propagation neural network can be applied in evaporative cooling air-conditioning system simulation, and the simulation results shows that the system performance is improved by 9.8 % the size of heat exchanger can be reduced by 25.3~43.7 % using R410A.