公共自行車系統（Bicycle-Sharing System, BSS）作為解決公共運輸第一和最後一哩路問題的關鍵，各國政府積極推動其發展並不斷增設新站點。然而，隨著站點增加，BSS的騎乘路線和調度變得複雜，而需要更多資源來優化調度策略和規劃騎乘環境建置。本研究提出一種基於旅運需求預測的旅次分佈長期預測模型，旨在有效預測新設站點建立後的旅次分佈並針對其安排高效的調度策略以及騎乘環境建置成本投入區域。
而在進行旅運需求預測中的旅次發生借車量與還車量的預測，本研究探討議題將牽涉到具備歷史票證資料的既有站點和欠缺歷史票證資料的新設站點，兩種站點形式，有鑑於時間序列資料分析法相對於利用變數推估在預測上有更高的精度，對於既有站點以時間序列資料分析為主提出了統計模型的差分整合移動平均自我迴歸模型(ARIMA)和季節性差分整合移動平均自我迴歸模型(SARIMA)和機器學習的長短期記憶(LSTM)和閥控遞迴單元(GRU)進行預測與比較，比較結果發現LSTM有最佳的預測精度；新設站點則利用氣候、人口、土地使用和周遭站點使用四個主要影響BSS借車量與還車量的因素，並應用統計模型的多元迴歸模型(MLR)和機器學習的極限梯度提升(XGBoost)進行預測和比較，XGBoost 有最佳的預測精度。因此本研究旅次發生階段則採納LSTM和XGBoost分別對既有站點和新設站點進行預測。旅次分佈則是應用傳統的重力模型與深度學習結合而成的深度重力(DG)模型，並以旅運需求預測程序為主要的概念，旅次發生的輸出結果加上距離作為旅次分佈的輸入，對BSS進行旅次分佈的預測，預測結果為0.7884，成功地捕捉到BSS旅次分佈樣態。
最後透過旅次分佈的結果配合新設站點的站點特性，發現若站點是以休閒觀光特性為主，則會以其他包含休閒光觀特性的站點為主要往來；住宅區則以前往休閒光觀景點、交通節點和就學為主要用途。

The Bicycle-Sharing System (BSS) is a key solution to the first and last mile problem in public transportation. Governments worldwide actively promote its development by continuously adding new stations. However, as the number of stations increases, BSS routes and dispatching become more complex, requiring more resources to optimize dispatch strategies and plan the riding environment. This study proposes a trip distribution prediction model based on travel demand prediction to effectively predict trip distribution after the establishment of new stations and to arrange efficient dispatch strategies and investment areas for the construction of the riding environment.
In predicting the number of borrowings and returns of trips in travel demand forecasting, this study explores the issues related to both existing stations with historical ticket data and new stations lacking this data. Given that time series analysis methods show higher accuracy than variable-based estimates, for existing stations, this study proposes the statistical models: the Autoregressive Integrated Moving Average (ARIMA) model, the Seasonal Autoregressive Integrated Moving Average (SARIMA) model, and machine learning models: the Long Short-Term Memory (LSTM) model and the Gated Recurrent Unit (GRU) model; and performs predictions and comparisons. Results indicate that LSTM has the best prediction accuracy. For new stations, the study utilizes four main factors that influence BSS borrowings and returns: climate, population, land use, and the usage of surrounding stations, and applies the statistical Multiple Linear Regression (MLR) model and the machine learning Extreme Gradient Boosting (XGBoost) model for predictions and comparisons, with XGBoost showing the best prediction accuracy. Therefore, this study adopts LSTM and XGBoost for predicting the trip generation of existing and new stations, respectively.For trip distribution, the study employs a Deep Gravity (DG) model, combining traditional gravity models with deep learning, using the travel demand prediction process as the main concept. The output results of trip generation, combined with distance, serve as inputs for trip distribution prediction. The prediction result has an R² value of 0.7884, accurately capturing the BSS trip distribution patterns.
Finally, through the trip distribution results and the characteristics of new stations, the study finds that if a station primarily serves recreational and tourism purposes, it will predominantly interact with other stations with similar characteristics. Residential areas mainly travel to recreational spots, transportation nodes, and educational institutions.

[1] 于妙善 ，2017，在需求與供應不確定下自行車共享系統多站點間之供需媒合，碩
士論文，臺灣大學工業工程學系，臺北市。
[2] 呂千慈，鍾易詩，馮正民，2018，環境空間與站點數對公共自行車租借量之影響
分析，碩士論文，國立交通大學運輸與物流管理學系，新竹市。
[3] 陳信宏，2022，基於階級關係之共享單車的調度策略，碩士論文，中原大學資訊
工程學系，桃園市。
[4] 陳惠國，2023，運輸規劃-基礎與進階，桃園市：五南。
[5] 陳惠國，2024，運輸工程，桃園市：五南。
[6] Azari, A., Papapetrou, P., Denic, S., Peters, G. (2019). Cellular traffic prediction and
classification: A comparative evaluation of LSTM and ARIMA. In Discovery Science:
22nd International Conference, DS 2019, Split, Croatia, October 28–30, 2019,
Proceedings 22 ,129-144. Springer International Publishing. doi:
https://doi.org/10.1007/978-3-030-33778-0_11
[7] Agresti, A. (2015). Foundations of linear and generalized linear models.
[8] Braun, L. M., Rodriguez, D. A., Cole-Hunter, T., Ambros, A., Donaire-Gonzalez, D.,
Jerrett, M., Nazelle, A. (2016). Short-term planning and policy interventions to promote
cycling in urban centers: Findings from a commute mode choice analysis in Barcelona,
Spain. Transportation Research Part A: Policy and Practice, 89, 164-183.
[9] Collini, E., Nesi, P., Pantaleo, G. (2021). Deep learning for short-term prediction of
available bikes on bike-sharing stations. IEEE Access, 9, 124337-124347. doi:
10.1109/ACCESS.2021.3110794.
[10] Chung, J., Gulcehre, C., Cho, K., Bengio, Y. (2014). Empirical evaluation of gated
recurrent neural networks on sequence modeling. arXiv preprint arXiv:1412.3555. doi:
https://doi.org/10.48550/arXiv.1412.3555
42
[11] Chen, T., Guestrin, C. (2016). Xgboost: A scalable tree boosting system. In Proceedings
of the 22nd acm sigkdd international conference on knowledge discovery and data
mining , 785-794. doi: https://doi.org/10.1145/2939672.2939785
[12] Chen, L. T., Hsu, Y. W. (2020). Socio-ecological predictors of frequent bike share trips:
Do purposes matter?. International Journal of Environmental Research and Public
Health, 17(20), 7640. Doi: https://www.mdpi.com/1660-4601/17/20/7640
[13] Feng, Y., Wang, S. (2017). A forecast for bicycle rental demand based on random forests
and multiple linear regression. In 2017 IEEE/ACIS 16th International Conference on
Computer and Information Science (ICIS) , 101-105. doi: 10.1109/ICIS.2017.7959977.
[14] Friedman, J. H. (2001). Greedy function approximation: a gradient boosting machine.
Annals of statistics, 1189-1232. doi: https://www.jstor.org/stable/2699986
[15] Guidon, S., Becker, H., Axhausen, K. (2019). Avoiding stranded bicycles in free-floating
bicycle-sharing systems: using survival analysis to derive operational rules for
rebalancing. 2019 IEEE Intelligent Transportation Systems Conference (ITSC), 1703
1708. IEEE.
[16] Guidon, S., Reck, D. J., Axhausen, K. (2020). Expanding a (n)(electric) bicycle-sharing
system to a new city: Prediction of demand with spatial regression and random forests.
Journal of Transport Geography, 84, 102692. doi:
https://doi.org/10.1016/j.jtrangeo.2020.102692
[17] Gu, X., Tang, X., Chen, T., Liu, X. (2024). Predicting the network shift of large urban
agglomerations in China using the deep-learning gravity model: A perspective of
population migration. Cities, 145, 104680. doi:
https://doi.org/10.1016/j.cities.2023.104680
[18] Habib, K. N., Mann, J., Mahmoud, M., & Weiss, A. (2014). Synopsis of bicycle demand
in the City of Toronto: Investigating the effects of perception, consciousness and
comfortability on the purpose of biking and bike ownership. Transportation research
43
part A: policy and practice, 70, 67-80.
[19] Hu, Z., Huang, K., Zhang, E., Ge, Q. A., Yang, X. (2021). Rebalancing strategy for bike
sharing systems based on the model of level of detail. Journal of advanced
transportation, 1-15. doi: https://doi.org/10.1155/2021/3790888
[20] Heanue, K. E., Pyers, C. E. (1966). A comparative evaluation of trip distribution
procedures. Highway Research Record, 114, 20-50.
[21] Hochreiter, S., Schmidhuber, J. (1997). Long short-term memory. Neural computation,
9(8), 1735-1780. doi: 10.1162/neco.1997.9.8.1735.
[22] Jiang, W. (2022). Bike sharing usage prediction with deep learning: a survey. Neural
Computing and Applications, 34(18), 15369-15385. doi: https://doi.org/10.1007/s00521
022-07380-5
[23] Liu, J., Li, Q., Qu, M., Chen, W., Yang, J., Xiong, H., Fu, Y. (2015). Station site
optimization in bike sharing systems. In 2015 IEEE International Conference on Data
Mining , 883-888. doi:10.1109/ICDM.2015.99
[24] Liang, Y., Ding, F., Huang, G., Zhao, Z. (2023). Deep trip generation with graph neural
networks for bike sharing system expansion. Transportation Research Part C: Emerging
Technologies, 154, 104241. doi: https://doi.org/10.1016/j.trc.2023.104241
[25] Lawson, H. C., Dearinger, J. A. (1967). A comparison of four work trip distribution
models. Journal of the Highway Division, 93(2), 1-25. doi:
https://doi.org/10.1061/JHCEA2.0000245
[26] Lenormand, M., Bassolas, A., Ramasco, J. J. (2016). Systematic comparison of trip
distribution laws and models. Journal of Transport Geography, 51, 158-169. doi:
https://doi.org/10.1016/j.jtrangeo.2015.12.008
[27] Lahoorpoor, B., Faroqi, H., Sadeghi-Niaraki, A., Choi, S. M. (2019). Spatial cluster
based model for static rebalancing bike sharing problem. Sustainability, 11(11), 3205.
doi: https://www.mdpi.com/2071-1050/11/11/3205
44
[28] Ma, X., Yin, Y., Jin, Y., He, M., Zhu, M. (2022). Short-term prediction of bike-sharing
demand using multi-source data: a spatial-temporal graph attentional LSTM approach.
Applied Sciences, 12(3), 1161.
[29] Mateo-Babiano, I., Bean, R., Corcoran, J., & Pojani, D. (2016). How does our natural
and built environment affect the use of bicycle sharing?. Transportation research part A:
policy and practice, 94, 295-307.
[30] Mateo-Babiano, I., Kumar, S., Mejia, A. (2017). Bicycle sharing in Asia: a stakeholder
perception and possible futures. Transportation research procedia, 25, 4966-4978. doi:
https://doi.org/10.1016/j.trpro.2017.05.375
[31] Midgley, P. (2011). Bicycle-sharing schemes: enhancing sustainable mobility in urban
areas. United Nations, Department of Economic and Social Affairs, 8, 1-12.
[32] Nagy, A. M., Simon, V. (2018). Survey on traffic prediction in smart cities. Pervasive
and Mobile Computing, 50, 148-163. doi: https://doi.org/10.1016/j.pmcj.2018.07.004
[33] Pan, Y., Zheng, R. C., Zhang, J., Yao, X. (2019). Predicting bike sharing demand using
recurrent neural networks. Procedia computer science, 147, 562-566. doi:
https://doi.org/10.1016/j.procs.2019.01.217
[34] Pourebrahim, N., Sultana, S., Thill, J. C., Mohanty, S. (2018). Enhancing trip distribution
prediction with twitter data: comparison of neural network and gravity models. In
Proceedings of the 2nd acm sigspatial international workshop on ai for geographic
knowledge discovery, 5-8. doi: https://doi.org/10.1145/3281548.3281555
[35] Pitombo, C. S., de Souza, A. D., Lindner, A. (2017). Comparing decision tree algorithms
to estimate intercity trip distribution. Transportation Research Part C: Emerging
Technologies, 77, 16-32. doi: https://doi.org/10.1016/j.trc.2017.01.009
[36] Rumelhart, D. E., Hinton, G. E., Williams, R. J. (1986). Learning representations by
back-propagating errors. nature, 323(6088), 533-536. doi:
https://www.nature.com/articles/323533a0
45
[37] Simini, F., Barlacchi, G., Luca, M., Pappalardo, L. (2021). A deep gravity model for
mobility flows generation. Nature communications, 12(1), 6576. doi:
https://doi.org/10.1038/s41467-021-26752-4
[38] Sohrabi, S., Paleti, R., Balan, L., Cetin, M. (2020). Real-time prediction of public bike
sharing system demand using generalized extreme value count model. Transportation
Research Part A: Policy and Practice, 133, 325-336.
[39] Song, J., Zhang, L., Qin, Z., Ramli, M. A. (2022). Spatiotemporal evolving patterns of
bike-share mobility networks and their associations with land-use conditions before and
after the COVID-19 outbreak. Physica A: Statistical Mechanics and its
Applications, 592, 126819.
[40] Subramanian, M., Cho, J., Veerappampalayam Easwaramoorthy, S., Murugesan, A., &
Chinnasamy, R. (2023). Enhancing Sustainable Transportation: AI-Driven Bike Demand
Forecasting in Smart Cities. Sustainability, 15(18), 13840.
[41] Sun, Y., Li, Y., Borozan, S., Wang, G., Qiu, J., Strbac, G. (2023). Battery swapping
dispatch for self-sustained highway energy system based on spatiotemporal deep
learning traffic flow prediction. IEEE Transactions on Industry Applications, 60(1). doi:
10.1109/TIA.2023.3321713.
[42] Sathishkumar, V. E., Park, J., Cho, Y. (2020). Using data mining techniques for bike
sharing demand prediction in metropolitan city. Computer Communications, 153, 353
366. doi: https://doi.org/10.1016/j.comcom.2020.02.007
[43] Sun, J., He, Y., Zhang, J. (2023). A Cluster-Then-Route Framework for Bike Rebalancing
in Free-Floating Bike-Sharing Systems. Sustainability, 15(22), 15994. doi:
https://www.mdpi.com/2071-1050/15/22/15994
[44] Tekouabou, S. C. K. (2021). Intelligent management of bike sharing in smart cities using
machine learning and Internet of Things. Sustainable Cities and Society, 67, 102702. doi:
https://doi.org/10.1016/j.scs.2020.102702
46
[45] Tillema, F., Van Zuilekom, K. M., Van Maarseveen, M. F. (2006). Comparison of neural
networks and gravity models in trip distribution. Computer‐Aided Civil and
Infrastructure Engineering, 21(2), 104-119. doi: https://doi.org/10.1111/j.1467
8667.2005.00421.x
[46] Vallez, C. M., Castro, M., Contreras, D. (2021). Challenges and opportunities in dock
based bike-sharing rebalancing: A systematic review. Sustainability, 13(4), 1829. doi:
https://doi.org/10.3390/su13041829
[47] VE, S., Cho, Y. (2024). Season wise bike sharing demand analysis using random forest
algorithm. Computational Intelligence, 40(1), e12287. doi:
https://doi.org/10.1111/coin.12287
[48] Wu, X., Lyu, C., Wang, Z., Liu, Z. (2019). Station-level hourly bike demand prediction
for dynamic repositioning in bike sharing systems. In Smart transportation systems
2019, 19-27. doi:https://doi.org/10.1007/978-981-13-8683-1_3
[49] Xu, J., Huang, E., Chen, C. H., Lee, L. H. (2015). Simulation optimization: A review and
exploration in the new era of cloud computing and big data. Asia-Pacific Journal of
Operational Research, 32(3), 1550019. doi:
https://doi.org/10.1142/S0217595915500190
[50] Yang, Y., Jia, B., Yan, X. Y., Chen, Y., Song, D., Zhi, D., Gao, Z. (2023). Estimating
intercity heavy truck mobility flows using the deep gravity framework. Transportation
Research Part E: Logistics and Transportation Review, 179, 103320. doi:
https://doi.org/10.1016/j.tre.2023.103320
[51] Yamak, P. T., Yujian, L., Gadosey, P. K. (2019). A comparison between arima, lstm, and
gru for time series forecasting. In Proceedings of the 2019 2nd international conference
on algorithms, computing and artificial intelligence, 49-55. doi:
https://doi.org/10.1145/3377713.3377722
[52] Yang, S., Yu, X., Zhou, Y. (2020). Lstm and gru neural network performance comparison
47
study: Taking yelp review dataset as an example. In 2020 International workshop on
electronic communication and artificial intelligence (IWECAI), 98-101. doi:
10.1109/IWECAI50956.2020.00027.
[53] Zhang, S. M., Su, X., Jiang, X. H., Chen, M. L., Wu, T. Y. (2019). A Traffic Prediction
Method of Bicycle-sharing based on Long and Short term Memory Network. J. Netw.
Intell., 4(2), 17-29.