2019-01-012024-05-13https://scholars.lib.ntu.edu.tw/handle/123456789/653105摘要：本子計畫之目的在於開發一套供空氣感測器專用之資料分析技術，使得感測器能正確檢測、辨識有害空氣微粒和有毒氣體，以實現空氣品質量化監測感測器開發之總計畫目標。本計畫將以氣體陣列檢測元件進行氣體檢測，量測不同種類氣體分子之響應，接著對數據資料做訊號處理，並開發資料分析之數學模型以量化分析不同氣體分子在不同濃度下之特徵，以將不同種類、不同濃度之氣體做區分。在模型開發完成後，將以另一組數據資料驗證模型的信效度。本子計畫開發的新型資料分析技術將結合多個不同特徵分析方法之優點，預期對不同有害空氣微粒和有毒氣體將有絕佳的辨識度，並能最佳化感測器之檢測性能包括其靈敏度、專一性和測量重複性。開發完成之資料分析技術預期將帶動空氣品質量化分析技術之發展，並能運用在其他領域的資料識別，如分子生醫影像、診斷用臨床超音波影像和臨床病理檢測、非破壞檢測等。<br> Abstract: The purpose of this project is to develop a data analysis technique for accurately and sensitively identify harmful atmospheric particulate matters that will be used by our quantitative air quality sensor. Firstly, we will use an array sensor to measure the responses of different kinds of gas molecules. Next, we will use signal processing techniques to analyze the data and develop a new mathematical model to quantify the characteristics of responses of different kinds of gas molecules in different concentrations, in order to design an algorithm for differentiating different molecules. Once the development of the model is complete, we will use another set of data to test the reliability and validity of the model. The novel data analysis technique developed in this project will share several advantages of different feature analysis methods, and is expected to provide excellent performance for detecting harmful atmospheric particulate matters as well as maximize the abilities of the sensor including its sensitivity, specificity and reliability. The new data analysis technique can help the development of quantitative air quality monitoring technology, and can be applied to other fields like molecular imaging, clinical diagnostic ultrasound imaging, clinical pathological examination and non-destructive testing.空氣感測器有害氣體資料量化分析訊號處理特徵分析air quality sensoratmospheric particulate mattersquantitative data analysissignal processingfeature analysis