細懸浮微粒(PM2.5)濃度分布會隨時間與地點改變，使不同通行路徑產生不同之累積污染暴露。微型感測器(以下簡稱微感器)可補充中央測站之空間觀測密度，惟其量測結果仍需經適當校正。本研究以中部空品區資料建置微感器校正模型與PM2.5空間推估模型，據此產製臺中市50 m空間解析度之PM2.5圖資，再將不同時間之濃度分布轉換為道路路段暴露成本，以行人移動情境進行低暴露路徑規劃與分析。另利用2023年全年逐日資料，比較暴露減少日數、路徑距離、時段、季節及區域之減量差異，並建置PM2.5圖資自動化產製流程。
微感器校正模型採用隨機森林迴歸(Random Forest Regressor, RFR)，測試集R2為0.87，RMSE及MAE分別為3.55及1.73 μg/m3。PM2.5空間推估整合校正後微感器與氣象、植生、土地利用、污染源、交通路網及時間變數，比較五種機器學習模型後，選定輕量梯度提升機(Light Gradient Boosting Machine, LightGBM)之迴歸模型(LGBMR)為最終模型。其測試集R2為0.83，RMSE及MAE分別為4.31及2.10 μg/m3，外部驗證R2為0.65。SHAP (SHapley Additive exPlanations)分析結果中，校正後微感器PM2.5濃度為模型中重要性最高之變數。
低暴露路徑規劃分析使用10,000組起訖點樣本，於2023年362個有效分析日8時與17時之PM2.5推估結果，相較於最短路徑，低暴露路徑平均距離增加18.2 m，平均繞行比例為0.61%，平均PM2.5濃度由14.77降至14.59 μg/m3。各組平均全年累積暴露減少量為156,145 μg/m3·s，平均全年暴露減少率為0.62%，平均暴露減少日數為162日。暴露減少量與減少率均隨路徑距離增加，8時與17時之平均效益相近，6月至8月之減量較少；區域比較則以城區最高、海線次之，屯區及山線較低。最佳減量案例全年累積暴露減少量達3,670,780 μg/m3·s，減少率為12.57%；特定時段僅增加40 m，平均PM2.5濃度由21.83降至14.32 μg/m3，暴露減少率為33.65%。
PM2.5長期與短期暴露與呼吸系統、心血管疾病及死亡風險相關，因此降低行人日常移動過程中之累積PM2.5暴露，可作為個人健康防護策略之一。本研究以全年逐日資料評估低暴露路徑之累積減量效益，並透過最佳減量案例說明，在特定污染分布與路網條件下，小幅改道亦可能產生明顯減量。其效益受起訖點位置、PM2.5時空分布及路網可替代性影響。當鄰近存在距離相近且濃度偏低之替代道路時，小幅改道即可降低累積暴露。

Fine particulate matter (PM2.5) concentrations vary over time and across nearby streets, creating route-dependent differences in cumulative exposure. This study developed an integrated framework combining microsensor calibration, Geo-AI-based PM2.5 spatial estimation, and low-exposure route optimization. Data from the Central Taiwan Air Quality Region were used to calibrate microsensors with a Random Forest Regressor (RFR) and to compare five machine learning models for spatial estimation. The selected Light Gradient Boosting Machine Regressor (LGBMR) generated time-specific 50-m-resolution PM2.5 maps for Taichung City. Predicted PM2.5 concentrations were assigned to road segments and used to calculate segment-level exposure costs. Route performance was evaluated at 8:00 and 17:00 on 362 valid days in 2023 using 10,000 stratified origin-destination (OD) pairs. The RFR achieved a test R2 of 0.87, while the LGBMR achieved a test R2 of 0.83 and an external-validation R2 of 0.65. Across all pairs, low-exposure routes added an average of 18.2 m, corresponding to a 0.61% detour, while reducing mean route concentration from 14.77 to 14.59 μg/m3. The mean annual cumulative exposure reduction was 156,145 μg/m3·s, with an exposure reduction rate of 0.62%. In the best-reduction case, annual cumulative exposure decreased by 3,670,780 μg/m3·s, corresponding to a 12.57% reduction. The benefits varied with route length, season, subregion, spatiotemporal patterns of PM2.5, and the availability of nearby alternative routes.

Alvarez-Pedrerol, M., Rivas, I., López-Vicente, M., Suades-González, E., Donaire-Gonzalez, D., Cirach, M., de Castro, M., Esnaola, M., Basagaña, X., Dadvand, P., Nieuwenhuijsen, M., & Sunyer, J. (2017). Impact of commuting exposure to traffic-related air pollution on cognitive development in children walking to school. Environmental Pollution, 231, 837–844. https://doi.org/10.1016/j.envpol.2017.08.075
Bagkis, E., Hassani, A., Schneider, P., DeSouza, P., Shetty, S., Kassandros, T., Salamalikis, V., Castell, N., Karatzas, K., Ahlawat, A., & Khan, J. (2025). Evolving trends in application of low-cost air quality sensor networks: challenges and future directions. npj Climate and Atmospheric Science, 8(1), 335. https://doi.org/10.1038/s41612-025-01216-4
Bo, Y., Chang, L.-y., Guo, C., Lin, C., Lau, A. K. H., Tam, T., & Lao, X. Q. (2021). Reduced ambient PM2.5, better lung function, and decreased risk of chronic obstructive pulmonary disease. Environment International, 156, 106706. https://doi.org/10.1016/j.envint.2021.106706
Bo, Y. C., Yu, T., Guo, C., Lin, C. C., Yang, H. T., Chang, L.-y. Y., Thomas, G. N., Tam, T., Lau, A. K. H., & Lao, X. Q. (2023). Cardiovascular Mortality, Habitual Exercise, and Particulate Matter 2.5 Exposure: A Longitudinal Cohort Study. American Journal of Preventive Medicine, 64(2), 250–258. https://doi.org/10.1016/j.amepre.2022.09.004
Breiman, L. (2001). Random Forests. Machine Learning, 45(1), 5–32. https://doi.org/10.1023/A:1010933404324
Briggs, D. J., Collins, S., Elliott, P., Fischer, P., Kingham, S., Lebret, E., Pryl, K., Van Reeuwijk, H., Smallbone, K., & Van Der Veen, A. (1997). Mapping urban air pollution using GIS: a regression-based approach. International Journal of Geographical Information Science, 11(7), 699–718. https://doi.org/10.1080/136588197242158
Chen, L.-J., Ho, Y.-H., Lee, H.-C., Wu, H.-C., Liu, H.-M., Hsieh, H.-H., Huang, Y.-T., & Lung, S.-C. C. (2017). An Open Framework for Participatory PM2.5 Monitoring in Smart Cities. IEEE Access, 5, 14441–14454. https://doi.org/10.1109/ACCESS.2017.2723919
Chen, T., & Guestrin, C. (2016). XGBoost: A Scalable Tree Boosting System Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, https://doi.org/10.1145/2939672.2939785
Chu, H.-J., Ali, M. Z., & He, Y.-C. (2020). Spatial calibration and PM2.5 mapping of low-cost air quality sensors. Scientific Reports, 10(1), 22079. https://doi.org/10.1038/s41598-020-79064-w
Dijkstra, E. W. (1959). A note on two problems in connexion with graphs. Numerische Mathematik, 1(1), 269–271. https://doi.org/10.1007/BF01386390
Eeftens, M., Beelen, R., de Hoogh, K., Bellander, T., Cesaroni, G., Cirach, M., Declercq, C., Dėdelė, A., Dons, E., de Nazelle, A., Dimakopoulou, K., Eriksen, K., Falq, G., Fischer, P., Galassi, C., Gražulevičienė, R., Heinrich, J., Hoffmann, B., Jerrett, M.,…Hoek, G. (2012). Development of Land Use Regression Models for PM2.5, PM2.5 Absorbance, PM10 and PMcoarse in 20 European Study Areas; Results of the ESCAPE Project. Environmental Science & Technology, 46(20), 11195–11205. https://doi.org/10.1021/es301948k
Friedman, J. H. (2001). Greedy Function Approximation: A Gradient Boosting Machine. The Annals of Statistics, 29(5), 1189–1232. https://doi.org/10.1214/aos/1013203451
Gao, L.-N., Tao, F., Ma, P.-L., Wang, C.-Y., Kong, W., Chen, W.-K., & Zhou, T. (2022). A short-distance healthy route planning approach. Journal of Transport & Health, 24, 101314. https://doi.org/10.1016/j.jth.2021.101314
Hatzopoulou, M., Weichenthal, S., Barreau, G., Goldberg, M., Farrell, W., Crouse, D., & Ross, N. (2013). A web-based route planning tool to reduce cyclists' exposures to traffic pollution: A case study in Montreal, Canada. Environmental Research, 123, 58–61. https://doi.org/10.1016/j.envres.2013.03.004
Health Effects Institute. (2025). State of Global Air 2025: A report on air pollution and its role in the world's leading causes of death. https://www.stateofglobalair.org/resources/report/state-global-air-report-2025
Jiang, P., Gao, C., Zhao, J., Li, F., Ou, C., Zhang, T., & Huang, S. (2024). An exploration of urban air health navigation system based on dynamic exposure risk forecast of ambient PM2.5. Environment International, 190, 108793. https://doi.org/10.1016/j.envint.2024.108793
Just, A. C., Arfer, K. B., Rush, J., Lyapustin, A., & Kloog, I. (2025). XIS-PM2.5: A daily spatiotemporal machine-learning model for PM2.5 in the contiguous United States. Environmental Research, 271, 120948. https://doi.org/10.1016/j.envres.2025.120948
Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., & Liu, T.-Y. (2017). LightGBM: A Highly Efficient Gradient Boosting Decision Tree https://proceedings.neurips.cc/paper_files/paper/2017/file/6449f44a102fde848669bdd9eb6b76fa-Paper.pdf
Lee, H.-Y., Hsu, C.-W., Ho, J.-T., Wong, P.-Y., Asri, A. K., Chen, C.-Y., Chin, W.-S., Guo, Y.-L., Hung, C.-H., Chen, S.-C., Cedeño-Laurent, J. G., & Wu, C.-D. (2026). Exposure-Aware commuting: Geo-AI-Driven route optimization to reduce NO2 in urban road networks. Sustainable Cities and Society, 144, 107403. https://doi.org/10.1016/j.scs.2026.107403
Lee, Y.-M., Lin, G.-Y., Le, T.-C., Hong, G.-H., Aggarwal, S. G., Yu, J.-Y., & Tsai, C.-J. (2024). Characterization of spatial-temporal distribution and microenvironment source contribution of PM2.5 concentrations using a low-cost sensor network with artificial neural network/kriging techniques. Environmental Research, 244, 117906. https://doi.org/10.1016/j.envres.2023.117906
Li, J., & Heap, A. D. (2008). A Review of Spatial Interpolation Methods for Environmental Scientists. (Geoscience Australia Record 2008/23). Canberra, ACT, Australia: Geoscience Australia
Lundberg, S. M., & Lee, S.-I. (2017). A Unified Approach to Interpreting Model Predictions Advances in Neural Information Processing Systems, https://proceedings.neurips.cc/paper_files/paper/2017/file/8a20a8621978632d76c43dfd28b67767-Paper.pdf
Lung, S.-C. C., Wang, W.-C. V., Wen, T.-Y. J., Liu, C.-H., & Hu, S.-C. (2020). A versatile low-cost sensing device for assessing PM2.5 spatiotemporal variation and quantifying source contribution. Science of The Total Environment, 716, 137145. https://doi.org/10.1016/j.scitotenv.2020.137145
Murtagh, E. M., Mair, J. L., Aguiar, E., Tudor-Locke, C., & Murphy, M. H. (2021). Outdoor Walking Speeds of Apparently Healthy Adults: A Systematic Review and Meta-analysis. Sports Medicine, 51(1), 125–141. https://doi.org/10.1007/s40279-020-01351-3
Oh, J., Hevia-Ramos, G., Ha, E., Hong, Y.-C., Kim, H., & Lim, Y.-H. (2025). A Systematic Review and Meta-Analysis on Long-Term Exposure to Particulate Matter and All-Cause and Cause-Specific Mortality in the Asia-Pacific States. Journal of Korean Medical Science, 40(27). https://doi.org/10.3346/jkms.2025.40.e156
Prokhorenkova, L., Gusev, G., Vorobev, A., Dorogush, A. V., & Gulin, A. (2018). CatBoost: unbiased boosting with categorical features https://proceedings.neurips.cc/paper_files/paper/2018/file/14491b756b3a51daac41c24863285549-Paper.pdf
Rafiepourgatabi, M., Woodward, A., Salmond, J. A., & Dirks, K. N. (2021). The Impact of Route Choice on Active Commuters' Exposure to Air Pollution: A Systematic Review [Systematic Review]. Frontiers in Sustainable Cities, Volume 2 - 2020. https://doi.org/10.3389/frsc.2020.565733
Shakerdonyavi, A., & Yeganeh, B. (2026). High-resolution PM2.5 mapping and evaluation in Tehran using mobile monitoring, land-use regression, and explainable hybrid models. Urban Climate, 65, 102773. https://doi.org/10.1016/j.uclim.2026.102773
Shukla, K., Kumar, P., Mann, G. S., & Khare, M. (2020). Mapping spatial distribution of particulate matter using Kriging and Inverse Distance Weighting at supersites of megacity Delhi. Sustainable Cities and Society, 54, 101997. https://doi.org/10.1016/j.scs.2019.101997
Tsai, Y.-G., Wang, J.-Y., Yang, K. D., Yang, H.-Y., Yeh, Y.-P., Chang, Y.-J., Lee, J. H., Wang, S.-L., Huang, S.-K., & Chan, C.-C. (2025). Long-term PM2.5 exposure impairs lung growth and increases airway inflammation in Taiwanese school children. ERJ Open Research, 11(4). https://doi.org/10.1183/23120541.00972-2024
Tzeng, S., Lai, C.-W., & Huang, H.-C. (2023). Spatially adaptive calibrations of airbox PM2.5 data. Biometrics, 79(4), 3637–3649. https://doi.org/10.1111/biom.13819
Wang, W.-C. V., Lung, S.-C. C., & Liu, C.-H. (2020). Application of machine learning for the in-field correction of a PM2.5 low-cost sensor network. Sensors, 20(17), 5002. https://doi.org/10.3390/s20175002
Wang, Y., Wu, Y., Li, Z., Liao, K., Li, C., & Song, G. (2022). Route planning for active travel considering air pollution exposure. Transportation Research Part D: Transport and Environment, 103, 103176. https://doi.org/10.1016/j.trd.2022.103176
Wong, P.-Y., Lee, H.-Y., Chen, Y.-C., Zeng, Y.-T., Chern, Y.-R., Chen, N.-T., Candice Lung, S.-C., Su, H.-J., & Wu, C.-D. (2021). Using a land use regression model with machine learning to estimate ground level PM2.5. Environmental Pollution, 277, 116846. https://doi.org/10.1016/j.envpol.2021.116846
Wong, P.-Y., Su, H.-J., Candice Lung, S.-C., Liu, W.-Y., Tseng, H.-T., Adamkiewicz, G., & Wu, C.-D. (2024). Explainable geospatial-artificial intelligence models for the estimation of PM2.5 concentration variation during commuting rush hours in Taiwan. Environmental Pollution, 349, 123974. https://doi.org/10.1016/j.envpol.2024.123974
World Health Organization. (2021). WHO global air quality guidelines: particulate matter (PM2.5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide. Geneva: World Health Organization Retrieved from https://www.who.int/publications/i/item/9789240034228
World Meteorological Organization. (2024). Integrating Low-cost Sensor Systems and Networks to Enhance Air Quality Applications. https://library.wmo.int/records/item/68924-integrating-low-cost-sensor-systems-and-networks-to-enhance-air-quality-applications?offset=1
Wu, C.-D., Chen, Y.-C., Pan, W.-C., Zeng, Y.-T., Chen, M.-J., Guo, Y. L., & Lung, S.-C. C. (2017). Land-use regression with long-term satellite-based greenness index and culture-specific sources to model PM2.5 spatial-temporal variability. Environmental Pollution, 224, 148–157. https://doi.org/10.1016/j.envpol.2017.01.074
Yen, J. Y. (1971). Finding the K Shortest Loopless Paths in a Network. Management Science, 17(11), 712–716. https://doi.org/10.1287/mnsc.17.11.712
Yu, W., Xu, R., Ye, T., Abramson, M. J., Morawska, L., Jalaludin, B., Johnston, F. H., Henderson, S. B., Knibbs, L. D., Morgan, G. G., Lavigne, E., Heyworth, J., Hales, S., Marks, G. B., Woodward, A., Bell, M. L., Samet, J. M., Song, J., Li, S., & Guo, Y. (2024). Estimates of global mortality burden associated with short-term exposure to fine particulate matter (PM2.5). The Lancet Planetary Health, 8(3), e146–e155. https://doi.org/10.1016/S2542-5196(24)00003-2
行政院主計總處. (2026). 115年1月1日常住人口統計結果. Retrieved from https://www.stat.gov.tw/News_Content.aspx?n=3703&s=235805
吳治達、曾于庭. (2023). 氣候變遷下的空氣污染分布：地理人工智慧技術之應用. 土木水利, 50(1), 16–23. https://doi.org/10.6653/MoCICHE.202302_50(1).0004
臺中市政府. (2026). 台中空品改善近三成 朝國家目標12微克邁進. Retrieved 2026/8/15 from https://www.taichung.gov.tw/8868/8872/9962/3240745
環境部. (2024). 空氣品質標準修正. Retrieved 2026/8/15 from https://air.moenv.gov.tw/envtopics/AirQuality_20.aspx
環境部. (2026a). 空氣污染物排放量清冊TEDS13.0版. Retrieved from https://air.moenv.gov.tw/EnvTopics/AirQuality_6.aspx
環境部. (2026b). 空氣品質與日常生活. Retrieved 2026/7/10 from https://airtw.moenv.gov.tw/cht/Encyclopedia/pedia08/pedia8_5.aspx
環境部. (2026c). 夏天的空氣品質為什麼比冬天好？. Retrieved 2026/7/13 from https://airtw.moenv.gov.tw/CHT/Encyclopedia/pedia08/pedia8_2.aspx
環境部. (2026d). 國家監測站、智慧城鄉感測點及學校民間感測器. Retrieved 2026/6/6 from https://airtw.moenv.gov.tw/cht/Encyclopedia/AirSensor/AirSensor_2.aspx
環境部. (2026e). 細懸浮微粒一日變化特徵. Retrieved 2026/7/2 from https://airtw.moenv.gov.tw/cht/Encyclopedia/pedia09/pedia9_3.aspx