類神經網路與PID混成控制器應用於空調系統溫度控制

The Artificial Neural Network and PID Hybrid Controller Applied to Temperature Control for Air Conditioning Systems

朱明輝(Ming-Huei Chu) ； 陳一維(Yi-Wei Cheng) ； 楊智勝(Chih-Sheng Yang) ； 廖展民(Jan-Min Liao)

《機械技師學刊》 8卷2期 (2015/08) Pp. 6-11

https://doi.org/10.6208/JPME-2015-06-02

摘要

空調系統為了達成節能功能，常採用變冷媒流量(VRF)空調系統，也就是市場所稱變頻空調系統。VRF空調系統，藉由改變壓縮機轉速改變冷媒流量，常應用於家用空調或小型辦工場所。現代VRF空調系統使用變頻壓縮機具有非線性特性且須適合不同安裝環境及變動負載等特性需要，傳統PID控制已不敷現代變頻空調系統須求，為了有效控制變頻空調系統，本研究採用類神經網路與PI混成控制應用於VRF空調系統，進行室內溫度控制。目的在強化適應性有效進行室內溫度控制及達成節能效果，模擬結果顯示類神經網路與PI混成控制器具有良好適應性，有效進行室內溫度控制，節能效果優於使用傳統PI控制的變頻空調系統。

關鍵字

變頻空調系統 ； 變冷媒流量 ； 類神經網路 ； 混成控制

並列摘要

The air conditioning systems apply variable refrigerant flow (VRF) techniques for Energy saving, which also called VRF air conditioning systems. The VRF air conditioning systems can change the refrigerant flow by changing the compressor speed, and usually applied to home and small office space. The modern VRF air conditioning systems applied variable speed compressors with nonlinear characteristics, and the varied room space and refrigerating load requirements lead the PID control becomes insufficient for modern VRF air conditioning systems. This study utilizes the neural network and PI hybrid controls applied to VRF air conditioner system for room temperature control, which can enhance the adaptability and saving energy. The simulation results reveal that the proposed control system can enhance the adaptability and be available with better energy saving performance than PI control.

並列關鍵字

Variable-frequency drive air conditioning systems ； Variable refrigerant flow ； Artificial neural networks ； Hybrid control