捷運場站使用手機之行人特性研究―基於社會力模型

Research on the Features of Pedestrians Using Smartphones at MRT stations―Based on Social Force Model

毛美能(Mei-Neng Mao)

指導教授 ： 許聿廷

國立臺灣大學/工學院/土木工程學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202102686

摘要

隨著智慧型手機的興起，邊走路邊使用手機的行為愈發普遍。因注意力無法集中於周遭環境變化，導致行人的感知能力下降，美國州際公路安全協會 (Governors Highway Safety Association, GHSA) 於2017年報告中表示行人的傷亡人數與使用手機的成長高度相關，且國發會調查 (2019) 接近37% 的臺灣受訪者曾有邊走路邊使用手機的經驗，然而目前社會對此行為並無太多的關注。使用手機在人流中會形成安全威脅，在捷運場站中更可能降低人流的紓解效率。本研究基於社會力模型，以數學模式描述並探討使用手機行為下捷運場站之行人流特性，透過擷取自實際影像行人之軌跡資料，校估模型參數，比較手機使用者和一般行人行為特性之差異；並將模型應用至微觀行人模擬軟體Viswalk，以了解在各情境假設下行人使用手機對於捷運場站之人流影響。
社會力模型以物理角度描述行人行為，其移動為吸引力與排斥力之合力。本研究假設於捷運場站之移動受自身驅動力 (driving force) 與他人排斥力 (repulsive force) 影響；為了方便獲取行人軌跡，以可提供由上往下拍攝角度的台北捷運忠孝復興站與臺電大樓站為拍攝場景，並藉由軌跡追蹤軟體Tracker擷取特定行人移動之時空資訊，針對有無使用手機的行人分別以基因演算法 (genetic algorithm) 校估模型之參數，以了解各參數的分布情形與相關性。
結果顯示使用手機者其反應時間相較於一般行人為長，且受外界斥力影響亦較大，容易形成移動速率、方向變化不穩定之情形；而在不同情境下的預測比較亦顯示，在複雜度較高的情境中，如人流密度高、無自組織行為 (self-organized behavior)，有無使用手機行人的特性差異將會擴大，意即使用手機的行為雖在密度低、情境較不複雜的情形下無太大影響，但在複雜情形下將與一般行人有明顯差異。將所得參數應用至模擬軟體Viswalk中，其結果顯示人流隨著行人總量與使用手機比例的增加，旅行時間將提升25%，且潛在衝突亦提高。針對研究發現，本研究提出對應建議，希冀能以此研究提升社會與政府對相關議題之關注，深化國人走路同時使用手機之危機意識。

關鍵字

使用智慧型手機之行人特性 ； 社會力模型 ； 軌跡投影 ； 基因演算法 ； Viswalk

並列摘要

With the increasing popularity of smartphones, the behavior of using smartphones while walking has become more common. The inability to focus on changes in the surroundings has led to a lower perception of pedestrians. Governors Highway Safety Association (GHSA) indicates that the number of pedestrian fatalities is highly correlated with the growth of phone usage. Additionally, nearly 37% of the Taiwanese respondents in the National Development Council survey (2019) revealed the experience of using smartphones while walking. However, society nowadays does not pay much attention to this issue. The use of mobile phones not only poses a threat to the safety of pedestrian flows; it may also reduce the efficiency of alleviating passenger flows at MRT stations. Based on the social force model, this study seeks to describe such behavior mathematically and exploring the characteristics of the pedestrians using smartphones at MRT stations. By extracting people’s trajectory data from the video filmed in the field, we calibrate the parameters in the model, and compares the behavioral difference between the pedestrians with or without using smartphones. Thereby, the results may be able to be further applied to the micro pedestrian simulation software, Viswalk, to examine the impact of smartphone users on the flow at MRT stations under various scenarios.
The social force model describes pedestrian behavior from a physical perspective, primarily considering the attractive and repulsive forces. It is assumed that the movement of people is determined by their driving forces and the repulsive forces received from others. To better capture people’s trajectory, we select the MRT stations allowing to take overhead shots: Zhongxiao Fuxing Station and Taipower Building Station. Tracker, a trajectory-capturing software is then applied to trace the movements, providing the position and time information of the target pedestrians at each interval, including normal pedestrians and those using smartphones.
The results show that pedestrians using smartphones need longer relaxation time; further, they are prone to be affected by the repulsive forces from the environment, suggesting that their speeds and directions be relatively unstable. The behavioral difference between normal pedestrians and smartphone users tends to expand in complex situations, such as the flows with high crowd density and without self-organization. Applying the obtained parameters to Viswalk, the simulation results demonstrate that travel time is uplifted by 25%, and potential conflicts also increase, with elevating flow volume and the proportion of smartphone users. Based om the findings of this research, we propose the suggestions to arouse the attention of the society and relevant authorities, hoping to raise the people’s awareness of the threat of walking while using smartphones.

並列關鍵字

Pedestrians Using Smartphones ； Social Force Model ； Trajectory Projection ； Genetic Algorithm ； Viswalk

參考文獻

Alsaleh, R., Sayed, T., Zaki, M. H. (2018). Assessing the effect of pedestrians’ use of smartphones on their walking behavior: a study based on automated video analysis. Transportation Research Record, 2672(35), 46-57.

Asano, M., Iryo, T., Kuwahara, M. (2010). Microscopic pedestrian simulation model combined with a tactical model for route choice behaviour. Transportation Research Part C: Emerging Technologies, 18(6), 842-855.

Association, G. H. S. (2017). Pedestrian Traffic Fatalities by State.

Benner, H., Kretz, T., Lohmiller, J., Sukennik, P. (2017). Is calibration a straight-forward task if detailed trajectory data is available? Paper presented at the Transportation Research Board 96th Annual Meeting.

Brown, D., Cox, A. J. (2009). Innovative uses of video analysis. The Physics Teacher, 47(3), 145-150.