賀耀華Ho, Yao-Hua朱家緯Chu, Ka-Ui2023-12-082023-01-012023-12-082022https://etds.lib.ntnu.edu.tw/thesis/detail/085780c0d83e305da762f9243d75e7fa/http://rportal.lib.ntnu.edu.tw/handle/20.500.12235/121550空氣汙染對人體會有負面的影響，空氣品質感測器 (空氣盒子) 可以定時紀錄室內空氣品質狀況，無論是狀態的改變、或是汙染物的累積，皆可以透過空氣盒子的數值來觀測。室外空氣品質的改善需透過眾人合作，而室內的空氣改善較能透過一己之力來達成。為了改善空氣品質，本研究希望對場域內不同的場地的空氣品質進行分群，基於此分群結果可以進行後續的分析與改善。本研究使用了位於台灣的校園室內空氣品質資料，在沒有任何其他外部資訊的條件下，單純使用每個場域的空氣品質數值的走勢，對該場域的空氣品質進行分群，當中使用了時間序列分解、快速傅立葉轉換 (fast Fourier transform, FFT)，以提升分群效率以及提取所需資訊。最後的結果顯示在沒有地理資訊或使用情形的條件下，分群結果可以反映出空間的通風狀況。Air pollution is a severe problem for the global environment. It is essential to improve air pollution. The air quality sensors (airbox) can record the air quality automatically, whether it is the change of status or the accumulation of pollutants. While outdoor air quality improvement requires the cooperation of many people, we can improve indoor air quality skillfully. This study aims to cluster the air quality into different clustering without other external information such as geographical location or field usage. Based on the results of this clustering, provide further analysis and improvement.This study uses indoor air quality data from the campus in Taiwan without any other external information. We apply K-means as the clustering strategy and use time-series decomposition and fast Fourier transform (FFT) to improve the efficiency of the clustering and extract the required feature. The final results show that without geographical information or usage conditions, the clustering results can reflect the ventilation of the space.NoneIndoor Air Quality ClusteringTime-series ClusteringK-means Clusteringetd