整合計算流體力學的室內溫度控制和熱舒適控制實驗

Abstract

節能減碳成為了各國發展和研究的主要議題之一，而冷氣機為一般家用和辦公建築中最耗能的設備，使得冷氣的控制成為值得探討與研究的主題。冷氣機提供人們室內的舒適環境，但如果一昧追求節能，可能將會損害舒適度，失去原本空調的目的，甚至降低生產力。為了控制室內的舒適環境，本研究主要分為兩個部分: (1) 為了冷氣機而整合計算流體力學和溫度控制以及 (2) 基於熱舒適度指標Predicted Mean Vote (PMV)的熱舒適控制實驗。 由於大部分空調控制的模擬都假設室內空氣是完美混合的。但此假設對於空間的溫度和空氣的流動提供較少的資訊。因此，在第一個部分，本研究利用計算流體力學建立房間空調的模型，並將此模型連接到控制器，在致冷的情況下進行溫度控制。本論文是採用自調式控制器，可以藉由遞迴式最小平方法監測受控廠的參數變化，並且根據控制法則和受控廠的參數即時調整控制器參數，以達到想要的控制特性。另外，模擬中加入隨時間變化的參考輸入和干擾，以探討控制的性能，干擾包括燈具造成的熱負載變化和溫度變化。模擬結果顯示自調式控制器可以順利追蹤想要的溫度，並且能補償干擾的影響。 第二部分是利用基於熱舒適度指標PMV，在夏天致冷的情況下進行室內熱舒適度控制。本研究的控制架構是透過反求PMV模型，決定熱舒適的溫度以作為控制目標，並且利用溫度控制器調整冷氣的溫度設定，以滿足室內熱舒適的需求，而且使用無線網路來收集環境資訊和傳輸控制命令。藉此方法，不需要拆卸冷氣的內部裝置。本研究分別採用自適應性網路模糊推論系統和粒子演算法，以解決非線性多變數的逆PMV模型。控制的方法分別採用了前饋控制結合比例-積分-微分控制器，前饋控制結合模糊控制器，以及自調式控制器。與傳統的固定冷氣設定的方法比較，實驗結果顯示本研究的控制方法能夠將PMV維持在±0.5的舒適範圍內，並且能節省30 %以上的耗能。

Energy conservation becomes a main topic every country studies and develops. An air conditioner is the most energy-consuming equipment in offices and homes. Therefore, air conditioning control deserves study. Air conditioners provide comfort environment for people. However, excessive pursuit of energy saving will lose the original purpose of air conditioning and even decrease productivity. In order to control indoor environment, this study includes two parts: (1) integration of computational fluid dynamics (CFD) and temperature controllers for air-conditioner control and (2) thermal comfort control experiments based on predicted mean vote (PMV). Most literature assumes that indoor is perfectly mixed for air conditioning control simulations, but this assumption provides little information on spatial temperature and air flow. Therefore, in first part, this study used CFD method to establish a thermal model of a room and links the model to a controller for temperature control under cooling condition. This thesis adopts self-tuning control. The controller can monitor the changes in parameters of a plant by a recursive least square method and tunes the control parameters in real-time according to control law and plant parameters to achieve desire control performance. Furthermore, a time-varying reference input and disturbances including heat load changes due to lighting heat and temperature variations are exerted in the simulation to investigate the control performance. Simulation results show that the self-tuning controllers can successfully track the desired temperature and compensate the effects of disturbances. The second part conducts indoor thermal comfort control based on a thermal comfort index PMV under summer cooling condition. An inverse PMV model is used to determine a thermal comfort temperature as a control target and temperature controllers regulate the temperature setting of an air conditioner. A wireless network is used to collect environmental information and transmit the control command. In this control structure, interior devices of an air conditioner do not need to be disassembled. This study adopts an adaptive-network-based fuzzy inference system and a particle swarm algorithm to solve a nonlinear and multivariable inverse PMV model, respectively. The control methods are feedforward with proportional-integral-derivative feedback, feedforward with fuzzy feedback, and self-tuning controllers. Compared with conventional fixed temperature settings, the present control methods effectively maintain the PMV value within the thermal comfort range of ±0.5 and energy in this study is saved more than 30%.