Pub Date : 2022-03-09DOI: 10.1080/19439962.2022.2048762
Xinyuan Wang, Jian Li, Rongjie Yu
Abstract Subways are the backbone of many urban transportation systems in large cities around the world. However, disruptions of subway services, such as power-supply failure or signal failures, can cause severe travel delays due to the large carrying capacities of subway trains. Accurate predictions of the disruption durations of subway services are essential for emergency response. To predict the disruption durations of subway services, previous studies primarily used parametric models (e.g., the accelerated failure time (AFT) model), and less attention has been given to machine-learning models with high potential prediction ability and fewer parameter restrictions. This paper proposes a model to predict and explore factors that affect the disruption durations of subway services in Shanghai using machine-learning models. The longitudinal data released by the subway operator from 2012 to 2021 were collected and analyzed in this study. A random survival forest (RSF) was used to describe the disruption durations of the subway service, and influential factors, such as incident reason, incident occurrence time, and line-related variables were considered in the model. Results show that the RSF model (C-Index = 0.672) achieved better prediction accuracy than the traditional AFT model (the best C-Index = 0.609) based on the collected data in Shanghai. In addition, results indicate that incident reason, disruption location, and the time of disruption factors can significantly affect subway service disruption durations. The proposed model can be used as a tool to predict the disruption durations of subway service for better disruption management of the subway system.
{"title":"Modeling disruption durations of subway service via random survival forests: The case of Shanghai","authors":"Xinyuan Wang, Jian Li, Rongjie Yu","doi":"10.1080/19439962.2022.2048762","DOIUrl":"https://doi.org/10.1080/19439962.2022.2048762","url":null,"abstract":"Abstract Subways are the backbone of many urban transportation systems in large cities around the world. However, disruptions of subway services, such as power-supply failure or signal failures, can cause severe travel delays due to the large carrying capacities of subway trains. Accurate predictions of the disruption durations of subway services are essential for emergency response. To predict the disruption durations of subway services, previous studies primarily used parametric models (e.g., the accelerated failure time (AFT) model), and less attention has been given to machine-learning models with high potential prediction ability and fewer parameter restrictions. This paper proposes a model to predict and explore factors that affect the disruption durations of subway services in Shanghai using machine-learning models. The longitudinal data released by the subway operator from 2012 to 2021 were collected and analyzed in this study. A random survival forest (RSF) was used to describe the disruption durations of the subway service, and influential factors, such as incident reason, incident occurrence time, and line-related variables were considered in the model. Results show that the RSF model (C-Index = 0.672) achieved better prediction accuracy than the traditional AFT model (the best C-Index = 0.609) based on the collected data in Shanghai. In addition, results indicate that incident reason, disruption location, and the time of disruption factors can significantly affect subway service disruption durations. The proposed model can be used as a tool to predict the disruption durations of subway service for better disruption management of the subway system.","PeriodicalId":46672,"journal":{"name":"Journal of Transportation Safety & Security","volume":"84 1","pages":"215 - 237"},"PeriodicalIF":2.6,"publicationDate":"2022-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89682233","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-02-10DOI: 10.1080/19439962.2022.2033899
Md Julfiker Hossain, J. Ivan, Shanshan Zhao, Kai Wang, Sadia Sharmin, N. Ravishanker, Eric D. Jackson
Abstract The injury severity of a driver in a crash is significantly related to the driver’s age and gender and vehicle characteristics. Previous studies have used only information about the most severely injured driver to represent the crash severity, ignoring other drivers involved in the crash, which can also be important to explain the crash severity. This study uses demographic information of all drivers involved in a multi-vehicle crash to predict the injury severity of the most severely injured driver using a partial proportional odds model. Models incorporating demographic information and vehicle characteristics of all drivers and vehicles involved in a crash were compared with models considering only information about the most severely injured driver in terms of significance of factors and prediction accuracy. The results indicate that although young drivers are likely to have lower levels of injury severity compared to working-age drivers, injury severity increases if the proportion of young drivers increases in a multi-vehicle crash. Drivers indicated to be not at fault frequently were more severely injured than drivers at fault. Finally, the inclusion of all drivers’ demographic information shows an improvement in the prediction accuracy of crash severity of the most severely injured driver.
{"title":"Considering demographics of other involved drivers in predicting the highest driver injury severity in multi-vehicle crashes on rural two-lane roads in California","authors":"Md Julfiker Hossain, J. Ivan, Shanshan Zhao, Kai Wang, Sadia Sharmin, N. Ravishanker, Eric D. Jackson","doi":"10.1080/19439962.2022.2033899","DOIUrl":"https://doi.org/10.1080/19439962.2022.2033899","url":null,"abstract":"Abstract The injury severity of a driver in a crash is significantly related to the driver’s age and gender and vehicle characteristics. Previous studies have used only information about the most severely injured driver to represent the crash severity, ignoring other drivers involved in the crash, which can also be important to explain the crash severity. This study uses demographic information of all drivers involved in a multi-vehicle crash to predict the injury severity of the most severely injured driver using a partial proportional odds model. Models incorporating demographic information and vehicle characteristics of all drivers and vehicles involved in a crash were compared with models considering only information about the most severely injured driver in terms of significance of factors and prediction accuracy. The results indicate that although young drivers are likely to have lower levels of injury severity compared to working-age drivers, injury severity increases if the proportion of young drivers increases in a multi-vehicle crash. Drivers indicated to be not at fault frequently were more severely injured than drivers at fault. Finally, the inclusion of all drivers’ demographic information shows an improvement in the prediction accuracy of crash severity of the most severely injured driver.","PeriodicalId":46672,"journal":{"name":"Journal of Transportation Safety & Security","volume":"8 1","pages":"43 - 58"},"PeriodicalIF":2.6,"publicationDate":"2022-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85638056","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-02-07DOI: 10.1080/19439962.2022.2033900
Daiquan Xiao, Željko Šarić, X. Xu, Q. Yuan
Abstract In recent years the pedestrian deaths have been declining, but the pedestrian–vehicle death rate in Croatia is still pretty high. This study intended to investigate the injury severity of pedestrian–vehicle crashes and identify the influencing factors. To achieve this goal, the dataset was firstly collected from Traffic Accident Database System maintained by the Ministry of the Interior, Republic of Croatia from 2015 to 2019, and then latent cluster analysis was employed to identify homogenous clusters from heterogeneous dataset. Based on the classified dataset, unbalanced panel mixed ordered probit model was proposed. By analyzing the classes with different vehicles, the proposed model revealed a more complete understanding of significant variables and showed beneficial performance from the goodness-of-fit, while capturing the impact of exogenous variables to vary among different places, as well as accommodating the heterogeneity issue due to unobserved effects. Findings revealed that the proposed model can be considered as an alternative to determine the factors of injury severity and to deal with the heterogeneity issue. The results may provide potential insights for reducing the injury severity of pedestrian-vehicle crashes.
{"title":"Investigating injury severity of pedestrian–vehicle crashes by integrating latent class cluster analysis and unbalanced panel mixed ordered probit model","authors":"Daiquan Xiao, Željko Šarić, X. Xu, Q. Yuan","doi":"10.1080/19439962.2022.2033900","DOIUrl":"https://doi.org/10.1080/19439962.2022.2033900","url":null,"abstract":"Abstract In recent years the pedestrian deaths have been declining, but the pedestrian–vehicle death rate in Croatia is still pretty high. This study intended to investigate the injury severity of pedestrian–vehicle crashes and identify the influencing factors. To achieve this goal, the dataset was firstly collected from Traffic Accident Database System maintained by the Ministry of the Interior, Republic of Croatia from 2015 to 2019, and then latent cluster analysis was employed to identify homogenous clusters from heterogeneous dataset. Based on the classified dataset, unbalanced panel mixed ordered probit model was proposed. By analyzing the classes with different vehicles, the proposed model revealed a more complete understanding of significant variables and showed beneficial performance from the goodness-of-fit, while capturing the impact of exogenous variables to vary among different places, as well as accommodating the heterogeneity issue due to unobserved effects. Findings revealed that the proposed model can be considered as an alternative to determine the factors of injury severity and to deal with the heterogeneity issue. The results may provide potential insights for reducing the injury severity of pedestrian-vehicle crashes.","PeriodicalId":46672,"journal":{"name":"Journal of Transportation Safety & Security","volume":"217 1","pages":"83 - 102"},"PeriodicalIF":2.6,"publicationDate":"2022-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89359148","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-02-02DOI: 10.1080/19439962.2022.2033901
Jaeyoung Lee, Yanqi Lian, M. Abdel-Aty, Suyi Mao, Qing Cai
Abstract Many states in the United States have passed the primary enforcement seat-belt law. Though there is strong evidence from previous studies that enhanced seat-belt enforcement interventions can substantially increase seat-belt use, thereby reducing fatalities. It is still necessary to evaluate the long-term effects of implementing the primary seat-belt law. In this study, changes in fatalities over time after the primary seat-belt law enactment are investigated using before-and-after study with the comparison group methods for fatality modification factors (FMFs). This study confirms that the number of adult fatalities without seat-belt has significantly decreased by 17.29%. Another key finding is that the fatality rates in states with a higher maximum fine amount are significantly lower than those with a lower one, however, the decrease in fatality trend is not as effective above about $100 fine. Implementing the primary seat-belt law is significantly effective in reducing fatalities without seat-belt in the long-term. Meanwhile, the relationship between fatalities reduction and the maximum fine amount is not positively linear related. It is imperative that states with the secondary seat-belt laws must reform their seat-belt laws to the primary seat-belt law. An appropriate fine amount can be determined to maximize the effectiveness of the primary seat-belt law.
{"title":"Long-term safety evaluation of the primary seat-belt law","authors":"Jaeyoung Lee, Yanqi Lian, M. Abdel-Aty, Suyi Mao, Qing Cai","doi":"10.1080/19439962.2022.2033901","DOIUrl":"https://doi.org/10.1080/19439962.2022.2033901","url":null,"abstract":"Abstract Many states in the United States have passed the primary enforcement seat-belt law. Though there is strong evidence from previous studies that enhanced seat-belt enforcement interventions can substantially increase seat-belt use, thereby reducing fatalities. It is still necessary to evaluate the long-term effects of implementing the primary seat-belt law. In this study, changes in fatalities over time after the primary seat-belt law enactment are investigated using before-and-after study with the comparison group methods for fatality modification factors (FMFs). This study confirms that the number of adult fatalities without seat-belt has significantly decreased by 17.29%. Another key finding is that the fatality rates in states with a higher maximum fine amount are significantly lower than those with a lower one, however, the decrease in fatality trend is not as effective above about $100 fine. Implementing the primary seat-belt law is significantly effective in reducing fatalities without seat-belt in the long-term. Meanwhile, the relationship between fatalities reduction and the maximum fine amount is not positively linear related. It is imperative that states with the secondary seat-belt laws must reform their seat-belt laws to the primary seat-belt law. An appropriate fine amount can be determined to maximize the effectiveness of the primary seat-belt law.","PeriodicalId":46672,"journal":{"name":"Journal of Transportation Safety & Security","volume":"33 1","pages":"1976 - 1996"},"PeriodicalIF":2.6,"publicationDate":"2022-02-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76901276","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract Quasi-induced exposure theory requires the clear-cut assignment of crash responsibility for individual crash-involved drivers. The assignment method based on the citation by police officers poses a concern that the citation would be issued due to the nonmoving violations rather than the driving actions that directly contribute to the crash. Thus, the objective of the study is to improve the accuracy of citation-based responsibility assignments. Binary logistic regression is employed to identify the factors affecting the citation decision of the police officers. An ensemble machine learning method that combines random forest, neural network, and extreme gradient boosting classifiers is established to allocate the crash responsibility. The findings include that (1) the police citation is closely related to the presence of hazardous driving behavior, but it can also be influenced by several factors such as driver age, drinking status, and the collision impact point of the vehicle; and (2) compared to the conventional models, the ensemble machine learning methods have better performance for crash responsibility assignment in terms of accuracy, Kappa coefficient, and area under the curve. The study serves to provide a reliable crash responsibility assignment approach to improve the accuracy of exposure estimation.
{"title":"An ensemble machine learning method for crash responsibility assignment in quasi-induced exposure theory","authors":"Guopeng Zhang, Ying Cai, Xinguo Jiang, Yingfei Fan, Yue Zhou, Jun Qian","doi":"10.1080/19439962.2022.2026543","DOIUrl":"https://doi.org/10.1080/19439962.2022.2026543","url":null,"abstract":"Abstract Quasi-induced exposure theory requires the clear-cut assignment of crash responsibility for individual crash-involved drivers. The assignment method based on the citation by police officers poses a concern that the citation would be issued due to the nonmoving violations rather than the driving actions that directly contribute to the crash. Thus, the objective of the study is to improve the accuracy of citation-based responsibility assignments. Binary logistic regression is employed to identify the factors affecting the citation decision of the police officers. An ensemble machine learning method that combines random forest, neural network, and extreme gradient boosting classifiers is established to allocate the crash responsibility. The findings include that (1) the police citation is closely related to the presence of hazardous driving behavior, but it can also be influenced by several factors such as driver age, drinking status, and the collision impact point of the vehicle; and (2) compared to the conventional models, the ensemble machine learning methods have better performance for crash responsibility assignment in terms of accuracy, Kappa coefficient, and area under the curve. The study serves to provide a reliable crash responsibility assignment approach to improve the accuracy of exposure estimation.","PeriodicalId":46672,"journal":{"name":"Journal of Transportation Safety & Security","volume":"20 1","pages":"24 - 42"},"PeriodicalIF":2.6,"publicationDate":"2022-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87896321","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-29DOI: 10.1080/19439962.2021.2020945
Mingjie Feng, Xuesong Wang, Yan Li
Abstract Freeways in China have developed rapidly in recent years. The large traffic volumes and high travel speeds have created a serious safety problem that is of growing concern, however. Accurate identification of factors influencing crashes is a prerequisite for implementing countermeasures, but unobserved heterogeneity in crash data can lead to erroneous inferences. To identify key factors influencing crash occurrence, this study used two data preparation and modeling approaches to account for unobserved heterogeneity. First, freeway traffic crashes were divided into single-vehicle (SV) and multi-vehicle (MV) crashes because of their different mechanisms of occurrence. Second, random parameter modeling and finite mixture modeling were used, and were compared with regard to their ability to account for unobserved heterogeneity. The results indicated that the finite mixture negative binomial regression model with two components (FMNB-2) produced a better goodness-of-fit and parameter estimation. Results of the FMNB-2 SV and MV models’ classification of crashes into two homogeneous subgroups showed that for both SV and MV, crashes in Component 1 were most affected by roadway geometric features, while in Component 2, crashes were more strongly associated with traffic operational conditions. These findings will help traffic managers implement more targeted countermeasures for freeway safety improvement.
{"title":"Analyzing single-vehicle and multi-vehicle freeway crashes with unobserved heterogeneity","authors":"Mingjie Feng, Xuesong Wang, Yan Li","doi":"10.1080/19439962.2021.2020945","DOIUrl":"https://doi.org/10.1080/19439962.2021.2020945","url":null,"abstract":"Abstract Freeways in China have developed rapidly in recent years. The large traffic volumes and high travel speeds have created a serious safety problem that is of growing concern, however. Accurate identification of factors influencing crashes is a prerequisite for implementing countermeasures, but unobserved heterogeneity in crash data can lead to erroneous inferences. To identify key factors influencing crash occurrence, this study used two data preparation and modeling approaches to account for unobserved heterogeneity. First, freeway traffic crashes were divided into single-vehicle (SV) and multi-vehicle (MV) crashes because of their different mechanisms of occurrence. Second, random parameter modeling and finite mixture modeling were used, and were compared with regard to their ability to account for unobserved heterogeneity. The results indicated that the finite mixture negative binomial regression model with two components (FMNB-2) produced a better goodness-of-fit and parameter estimation. Results of the FMNB-2 SV and MV models’ classification of crashes into two homogeneous subgroups showed that for both SV and MV, crashes in Component 1 were most affected by roadway geometric features, while in Component 2, crashes were more strongly associated with traffic operational conditions. These findings will help traffic managers implement more targeted countermeasures for freeway safety improvement.","PeriodicalId":46672,"journal":{"name":"Journal of Transportation Safety & Security","volume":"12 1","pages":"59 - 81"},"PeriodicalIF":2.6,"publicationDate":"2021-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82409143","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract Crash risk prediction plays a pivotal role in traffic management and infrastructure optimization. Previous research has studied the relationship between crashes and multiple factors using statistical methods. As both drivers’ attention and environmental complexity substantially affect traffic safety, this article presents a novel method to predict crash risk proactively by combining these two interactive factors. More than 200 high-risk zones and 300 noncrash zones were screened out through social media data. Corresponding environmental information was collected using the street view map. Spectral saliency mapping was applied to depict the driver’s attention distribution toward images. A DeepLabV3 pretrained network was implemented to label the semantic features in the environment. A featured vector was then constructed by fuzing the visual attention model and image semantics. The gradient boosting decision tree algorithm was applied to analyze the relationship between the multitype crash data and featured vectors. The results showed that the accuracy of the proposed method for detecting different types of crashes was over 0.81. Dynamic objects are the most substantial factors that affect crash possibility and categories. Traffic signals are vulnerable to drivers’ attention, which may be easily overlooked. The proposed method provides new insights into understanding traffic crash risk, which can help us predict different types of crashes more effectively.
{"title":"An image-based crash risk prediction model using visual attention mapping and a deep convolutional neural network","authors":"Chengyu Hu, Wenchen Yang, Chenglong Liu, Rui Fang, Zhongyin Guo, Bijiang Tian","doi":"10.1080/19439962.2021.2015731","DOIUrl":"https://doi.org/10.1080/19439962.2021.2015731","url":null,"abstract":"Abstract Crash risk prediction plays a pivotal role in traffic management and infrastructure optimization. Previous research has studied the relationship between crashes and multiple factors using statistical methods. As both drivers’ attention and environmental complexity substantially affect traffic safety, this article presents a novel method to predict crash risk proactively by combining these two interactive factors. More than 200 high-risk zones and 300 noncrash zones were screened out through social media data. Corresponding environmental information was collected using the street view map. Spectral saliency mapping was applied to depict the driver’s attention distribution toward images. A DeepLabV3 pretrained network was implemented to label the semantic features in the environment. A featured vector was then constructed by fuzing the visual attention model and image semantics. The gradient boosting decision tree algorithm was applied to analyze the relationship between the multitype crash data and featured vectors. The results showed that the accuracy of the proposed method for detecting different types of crashes was over 0.81. Dynamic objects are the most substantial factors that affect crash possibility and categories. Traffic signals are vulnerable to drivers’ attention, which may be easily overlooked. The proposed method provides new insights into understanding traffic crash risk, which can help us predict different types of crashes more effectively.","PeriodicalId":46672,"journal":{"name":"Journal of Transportation Safety & Security","volume":"20 1","pages":"1 - 23"},"PeriodicalIF":2.6,"publicationDate":"2021-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82606291","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-28DOI: 10.1080/19439962.2021.2005727
Wang Xiang, Li Chen, Bin Wang, Qingwan Xue, Wei Hao, Xuemei Liu
Abstract To secure the city against the transmission of COVID-19, metro ridership control is an important task of the metro corporation on the premise of meeting the basic travel demand as far as possible. First off, this paper describes the influence mechanism of COVID-19 on metro ridership in Changsha, including an analysis into the correlation among policies, population, and metro ridership. Secondly, this paper verifies the influence of governmental macro-policy on population mobility and metro train working diagram, and thereby on metro ridership, based on the actual data during Jan 12th to May 6th, 2020 and year-ago data (2019). And then the Difference-in-Difference (DID) model is used to verify the effect of policies on metro ridership in Changsha. Results also show the effectiveness of policy chain on the limit of metro ridership, which bears a strong correlation to the number of confirmed COVID-19 cases. A linear regression prediction model is built to predict metro ridership based on cumulative net inflow of population index, metro carrying capacity and confirmed COVID-19 cases. This paper is expected to provide reference for Metro Corporation to control ridership on the premise of meeting the basic travel demand amid the explosive outbreak of the epidemics.
{"title":"Policies, population and impacts in metro ridership response to COVID-19 in Changsha","authors":"Wang Xiang, Li Chen, Bin Wang, Qingwan Xue, Wei Hao, Xuemei Liu","doi":"10.1080/19439962.2021.2005727","DOIUrl":"https://doi.org/10.1080/19439962.2021.2005727","url":null,"abstract":"Abstract To secure the city against the transmission of COVID-19, metro ridership control is an important task of the metro corporation on the premise of meeting the basic travel demand as far as possible. First off, this paper describes the influence mechanism of COVID-19 on metro ridership in Changsha, including an analysis into the correlation among policies, population, and metro ridership. Secondly, this paper verifies the influence of governmental macro-policy on population mobility and metro train working diagram, and thereby on metro ridership, based on the actual data during Jan 12th to May 6th, 2020 and year-ago data (2019). And then the Difference-in-Difference (DID) model is used to verify the effect of policies on metro ridership in Changsha. Results also show the effectiveness of policy chain on the limit of metro ridership, which bears a strong correlation to the number of confirmed COVID-19 cases. A linear regression prediction model is built to predict metro ridership based on cumulative net inflow of population index, metro carrying capacity and confirmed COVID-19 cases. This paper is expected to provide reference for Metro Corporation to control ridership on the premise of meeting the basic travel demand amid the explosive outbreak of the epidemics.","PeriodicalId":46672,"journal":{"name":"Journal of Transportation Safety & Security","volume":"64 1","pages":"1955 - 1975"},"PeriodicalIF":2.6,"publicationDate":"2021-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89861072","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-10DOI: 10.1080/19439962.2021.2011810
Chang Peng, Chengcheng Xu
Abstract The primary objective of this paper is to develop a combined variable speed limit (VSL) and lane change guidance (LCG) controller to prevent secondary crashes (SCs) and improve traffic efficiency on freeways. VSL controllers deliver speed limit instructions and LCG controllers deliver lane-changing instructions. A distributed deep reinforcement learning (RL)–based combined controller was proposed. The performance of the combined controller was evaluated in terms of safety and efficiency. Simulation experiments indicated that due to the complementation of VSL and LCG, the developed combined controller achieved higher performance in general than any single subcontroller. VSL control in a combined controller contributed prior effects on SC prevention and efficiency improvement, while LCG control improved the drawback of VSL by reducing the number of tough lane changes and avoiding extra SC risks caused by speed limit in relatively uncongested conditions. Moreover, the results of attention area investigation and sensitivity analysis revealed that the developed controller was able to accurately capture the spatial and temporal impact areas caused by prior crashes and generate proper interventions of traffic flow proactively.
{"title":"Combined variable speed limit and lane change guidance for secondary crash prevention using distributed deep reinforcement learning","authors":"Chang Peng, Chengcheng Xu","doi":"10.1080/19439962.2021.2011810","DOIUrl":"https://doi.org/10.1080/19439962.2021.2011810","url":null,"abstract":"Abstract The primary objective of this paper is to develop a combined variable speed limit (VSL) and lane change guidance (LCG) controller to prevent secondary crashes (SCs) and improve traffic efficiency on freeways. VSL controllers deliver speed limit instructions and LCG controllers deliver lane-changing instructions. A distributed deep reinforcement learning (RL)–based combined controller was proposed. The performance of the combined controller was evaluated in terms of safety and efficiency. Simulation experiments indicated that due to the complementation of VSL and LCG, the developed combined controller achieved higher performance in general than any single subcontroller. VSL control in a combined controller contributed prior effects on SC prevention and efficiency improvement, while LCG control improved the drawback of VSL by reducing the number of tough lane changes and avoiding extra SC risks caused by speed limit in relatively uncongested conditions. Moreover, the results of attention area investigation and sensitivity analysis revealed that the developed controller was able to accurately capture the spatial and temporal impact areas caused by prior crashes and generate proper interventions of traffic flow proactively.","PeriodicalId":46672,"journal":{"name":"Journal of Transportation Safety & Security","volume":"30 1","pages":"2166 - 2191"},"PeriodicalIF":2.6,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80352253","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-18DOI: 10.1080/19439962.2021.1994682
Jingfeng Ma, Gang Ren, Haoxuan Fan, Shunchao Wang, Jingcai Yu
Abstract Traffic crashes involving vehicles are mainly caused by illegal driving behaviors. It is of paramount importance to mitigate traffic violation occurrences. This study positions itself to characterize the effects of contributing factors on traffic violation severity. Considering different traffic violation outcomes caused by various factors, this study selects 17 factors from the spatiotemporal, road-traffic, vehicle-driver, and environment characteristics based on 55,997 valid traffic violations. A model comparison as well as the elasticity for the optimal model (partial proportional odds model) is applied to facilitate the related interpretation. The results evidenced the significant roles of time of day, vehicle type, driver age, interference, road type, weather, lighting condition, and speed limit. The findings revealed that higher-grade roads, higher speed limits, lower visibility, more interference, and increasing traffic volumes are significantly associated with a reduction in the slight probabilities but an increase in the more severe probabilities. Older drivers with more experience are correlated with a substantial increase in the slight probabilities yet an obvious decrease in the mild probabilities. The findings could provide meaningful insights to prioritize effective related countermeasures.
{"title":"Determinants of traffic violations in China: A case-study with a partial proportional odds model","authors":"Jingfeng Ma, Gang Ren, Haoxuan Fan, Shunchao Wang, Jingcai Yu","doi":"10.1080/19439962.2021.1994682","DOIUrl":"https://doi.org/10.1080/19439962.2021.1994682","url":null,"abstract":"Abstract Traffic crashes involving vehicles are mainly caused by illegal driving behaviors. It is of paramount importance to mitigate traffic violation occurrences. This study positions itself to characterize the effects of contributing factors on traffic violation severity. Considering different traffic violation outcomes caused by various factors, this study selects 17 factors from the spatiotemporal, road-traffic, vehicle-driver, and environment characteristics based on 55,997 valid traffic violations. A model comparison as well as the elasticity for the optimal model (partial proportional odds model) is applied to facilitate the related interpretation. The results evidenced the significant roles of time of day, vehicle type, driver age, interference, road type, weather, lighting condition, and speed limit. The findings revealed that higher-grade roads, higher speed limits, lower visibility, more interference, and increasing traffic volumes are significantly associated with a reduction in the slight probabilities but an increase in the more severe probabilities. Older drivers with more experience are correlated with a substantial increase in the slight probabilities yet an obvious decrease in the mild probabilities. The findings could provide meaningful insights to prioritize effective related countermeasures.","PeriodicalId":46672,"journal":{"name":"Journal of Transportation Safety & Security","volume":"90 1","pages":"1934 - 1954"},"PeriodicalIF":2.6,"publicationDate":"2021-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79937583","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：