- 學位論文
微型氣體感測器之效能比對與校正方法建立之研究
Performance Comparison and Calibration of Low Cost Gas Sensors
摘要
在網路蓬勃發展、資訊流通迅速的現代社會，物聯網(Internat of Things，IoT)成為生活中不可或缺的物件，同時人們對生活環境品質也越趨要求，因此建構次世代空品監測網(Next Generation of Air Monitoring，NGAM)，在城市中廣佈微型感測器，可便於民眾即時掌握周遭環境資訊。微型感測器優點在於低價、建置簡單以及容易取得；缺點在於其準確性與精密度較差，故本研究將建立合宜之微型氣體感測器之校正方法，並使用多元線性迴歸模型來考量環境變數，進而提升感測器之精確度。研究中使用電化學感測器及半導體感測器，對典型大氣濃度下的污染氣體：一氧化碳、一氧化氮、二氧化氮、臭氧以及總碳氫化合物進行監測與校正比對。在實驗過程中皆以FEM儀器同時平行比對，並對感測器之表現性進行探討。實驗室環境中，未校正之感測器與參考儀器間的一般線性迴歸結果為中度相關(R2>0.49)；經考量溫、濕度並以多元迴歸對感測器校正後，感測器與參考儀器間的相關程度提升至高度相關(R2>0.69)。除了校正之外，本研究利用準確度及精密度來檢視感測器的QA/QC是否符合期待，研究結果顯示感測器精密度除了TVOC感測器為70%與B系列NO2感測器為77%外，其餘感測器皆達90%以上；準確度部分則在使用多元迴歸校正後，依不同的感測器可達10至30%的提升，驗證感測器經多元校正後可提升其準確度及精密度。此外，本研究亦對野外環境進行感測器應用分析，經分析後得知感測器在野外易受風速影響，特別是NO2感測器，其受影響之權重甚至達50%以上，因此需考量野外環境因子並納入校正式。經由野外測試結果顯示，目前僅有CO感測器是有能力對野外大氣環境進行有效監測，其餘氣體感測器僅能用於警示較高濃度污染使用。
並列摘要
The Internet of Things (IoT) has become an indispensable part of modern life. In Taiwan, the increasingly demand on air quality calls for the Next Generation of Air Monitoring (NGAM) to be constructed. This can assist the public for easy access to the air quality of surrounding environment via the low cost sensors. In addition to the advantage of low cost, the air quality sensors can be easily assembled and distributed. However, their disadvantages include the low accuracy, low precision and less life time usage. Hence the purpose of this study is to establish an appropriate calibration procedure for commercialized low cost electrochemical and semiconductor sensors which can detect CO, NO, NO2, O3 and THC, respectively. The US EPA-approved FEM instruments were used for the standards of parallel comparison. Electrochemical sensors and semiconductor sensors were used in the present study. In the lab, the coefficient of determination (R2) of linear regression (LR) between the non-corrected sensors and the reference instrument are moderately correlated (R2> 0.49). Incorporating the temperature and humidity into the multiple linear regression (MLR) can help to increase the R2 to be larger than 0.69. In addition, this study checked the accuracy and precision of the gas sensors. Except for the precision of SGP-30 (73%) and NO2-B43F (77%), other sensors have acceptable precision of over 90%. It is also worth noting that the accuracy of gas sensors is significantly higher (increased by 10-30%) as calibrated by MLR method. Furthermore, the field MLR calibration results revealed that the wind speed has the highest weighting of all evaluated variables. For NO2 sensors, the weighting is especially high (> 50%). Therefore, the field environmental factors must be considered into the MLR calibration method, which can make sensors performed better. Afterwards, this study indicated that for the field monitoring application, only CO sensors can effectively monitor the ambient concentration; other gas sensors can only be used for warning high pollution status.
參考文獻
延伸閱讀
- Ting, G. C. (2022). 分析探討及實驗驗證精密微型軸流式氣渦輪之設計與操作參數對輸出性能影響關係 [doctoral dissertation, National Taiwan University]. Airiti Library. https://doi.org/10.6342/NTU202201994
- 林建谷（2004）。高準確性微壓力感測器之研究〔碩士論文，國立臺灣大學〕。華藝線上圖書館。https://doi.org/10.6342/NTU.2004.01465
- 李宇修（2009）。微型氣壓振動器之設計開發與性能研究〔碩士論文，國立臺灣大學〕。華藝線上圖書館。https://doi.org/10.6342/NTU.2009.02516
- 羅仁聰、吳志榮、黃財旺、陳中邦、郭景宜、何宜霖、蕭俊豪（2015）。High Pressure Gas Flow Calibration Standard Traceability and Comparison。石油季刊，51(4)，123-138。https://www.airitilibrary.com/Article/Detail?DocID=10229671-201512-201512180009-201512180009-123-138
- 楊力儼、柯廷勳、曾文甲（2019）。Introduction of Solid-state Gas Sensor。科儀新知，(218)，12-25。https://www.airitilibrary.com/Article/Detail?DocID=10195440-201903-201904080012-201904080012-12-25