Pub Date : 2024-10-20DOI: 10.1080/19427867.2023.2277013
This study introduces a novel parcel locker location problem in the last mile delivery where each parcel locker is equipped with a set of selectable volume sizes and provides services for two types of customers: commuters and home-based customers. We establish a pickup cost model to quantify the effort exerted by customers during the retrieval process. Furthermore, we design a satisfaction function to measure customers’ willingness to retrieve items from a specific parcel locker. Finally, we implement a financial incentive mechanism to enhance customer satisfaction by reducing the pickup cost. To solve this problem, we formulate a problem-specific scatter search algorithm. Extensive experiments demonstrate the computational effectiveness of the proposed algorithm. Furthermore, sensitivity analyses show that the selectable volume size model can reduce the total cost by up to 16.8%, while “pickup on the way” can achieve a reduction of up to 63.97% in our instances.
{"title":"Parcel locker location problem with selectable volume sizes and heterogeneous customers in the last mile delivery","authors":"","doi":"10.1080/19427867.2023.2277013","DOIUrl":"10.1080/19427867.2023.2277013","url":null,"abstract":"<div><div>This study introduces a novel parcel locker location problem in the last mile delivery where each parcel locker is equipped with a set of selectable volume sizes and provides services for two types of customers: commuters and home-based customers. We establish a pickup cost model to quantify the effort exerted by customers during the retrieval process. Furthermore, we design a satisfaction function to measure customers’ willingness to retrieve items from a specific parcel locker. Finally, we implement a financial incentive mechanism to enhance customer satisfaction by reducing the pickup cost. To solve this problem, we formulate a problem-specific scatter search algorithm. Extensive experiments demonstrate the computational effectiveness of the proposed algorithm. Furthermore, sensitivity analyses show that the selectable volume size model can reduce the total cost by up to 16.8%, while “pickup on the way” can achieve a reduction of up to 63.97% in our instances.</div></div>","PeriodicalId":48974,"journal":{"name":"Transportation Letters-The International Journal of Transportation Research","volume":"16 9","pages":"Pages 1140-1154"},"PeriodicalIF":3.3,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135286122","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 : 2024-10-20DOI: 10.1080/19427867.2023.2270243
In urban metro systems, unexpected events often occur to cause discordance in the planned timetable and delay in passenger travel. Therefore, this study focuses on the timetable rescheduling problem by adopting a stop-skipping strategy during the off-peak period. In addition, there are a number of existing optimization models that minimize the total passenger waiting time, which can cause fairness issues for a small group of passengers having longer waiting time at some stations. In this study, a rescheduling model is proposed to consider passenger waiting time fairness and train travel time cost. This model is tested with a randomly designed delay scenario of Nanjing Metro Line 2, which can be efficiently solved by a designed iterative algorithm based on decomposition method. Compared to the standard stop pattern, the result shows that the stop-skipping pattern can better alleviate the impact of sudden delays, while simultaneously ensuring the passenger waiting time fairness.
{"title":"Metro stop-skipping rescheduling optimization considering passenger waiting time fairness","authors":"","doi":"10.1080/19427867.2023.2270243","DOIUrl":"10.1080/19427867.2023.2270243","url":null,"abstract":"<div><div>In urban metro systems, unexpected events often occur to cause discordance in the planned timetable and delay in passenger travel. Therefore, this study focuses on the timetable rescheduling problem by adopting a stop-skipping strategy during the off-peak period. In addition, there are a number of existing optimization models that minimize the total passenger waiting time, which can cause fairness issues for a small group of passengers having longer waiting time at some stations. In this study, a rescheduling model is proposed to consider passenger waiting time fairness and train travel time cost. This model is tested with a randomly designed delay scenario of Nanjing Metro Line 2, which can be efficiently solved by a designed iterative algorithm based on decomposition method. Compared to the standard stop pattern, the result shows that the stop-skipping pattern can better alleviate the impact of sudden delays, while simultaneously ensuring the passenger waiting time fairness.</div></div>","PeriodicalId":48974,"journal":{"name":"Transportation Letters-The International Journal of Transportation Research","volume":"16 9","pages":"Pages 1115-1125"},"PeriodicalIF":3.3,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135888859","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 : 2024-10-20DOI: 10.1080/19427867.2023.2278854
Existing unconventional intersection designs can enhance the capacity of intersections to deal with repaid pressure on urban commuting, whereas they force vehicles to multi-stop and reorganize, resulting in safety risks and high fuel consumption. This paper proposes the Intermittent Lane Intersection (ILI) that combines the advantages of variable lane intersection and CTE (Combining the tandem control and exit lanes for left-turn control). An integrated optimization model is established to obtain the optimal geometric and signal control parameters, and a homothermic genetic algorithm is used to solve the model. Then, the application of ILI in different traffic demands of the Tianfu Third Street intersection is analyzed. Compared with the benchmark design (TI, CTE, and CTE with signal coordinated control), ILI can effectively reduce intersection delay, fuel consumption, and parking times. Additionally, ILI has great application superiority when the traffic arrival rate is [1500,1700] and the left-turn ratio is [0.40,0.60].
现有的非常规交叉口设计可以提高交叉口的通行能力,以应对城市通勤的补给压力,但却迫使车辆多次停车和重组,造成安全隐患和高油耗。本文提出的间歇车道交叉口(ILI)结合了可变车道交叉口和 CTE(结合串联控制和出口车道进行左转控制)的优点。建立了一个综合优化模型,以获得最佳几何和信号控制参数,并使用同温遗传算法对模型进行求解。然后,分析了 ILI 在天府三街交叉口不同交通需求下的应用。与基准设计(TI、CTE 和带信号协调控制的 CTE)相比,ILI 可有效减少交叉口延误、油耗和停车时间。此外,当交通到达率为[1500,1700],左转率为[0.40,0.60]时,ILI 具有很大的应用优势。
{"title":"The Integrated optimization of intermittent lane intersection design and dynamic signal control: efficiency, safety, and fuel consumption","authors":"","doi":"10.1080/19427867.2023.2278854","DOIUrl":"10.1080/19427867.2023.2278854","url":null,"abstract":"<div><div>Existing unconventional intersection designs can enhance the capacity of intersections to deal with repaid pressure on urban commuting, whereas they force vehicles to multi-stop and reorganize, resulting in safety risks and high fuel consumption. This paper proposes the Intermittent Lane Intersection (ILI) that combines the advantages of variable lane intersection and CTE (Combining the tandem control and exit lanes for left-turn control). An integrated optimization model is established to obtain the optimal geometric and signal control parameters, and a homothermic genetic algorithm is used to solve the model. Then, the application of ILI in different traffic demands of the Tianfu Third Street intersection is analyzed. Compared with the benchmark design (TI, CTE, and CTE with signal coordinated control), ILI can effectively reduce intersection delay, fuel consumption, and parking times. Additionally, ILI has great application superiority when the traffic arrival rate is [1500,1700] and the left-turn ratio is [0.40,0.60].</div></div>","PeriodicalId":48974,"journal":{"name":"Transportation Letters-The International Journal of Transportation Research","volume":"16 9","pages":"Pages 1155-1169"},"PeriodicalIF":3.3,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135242091","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 : 2024-10-20DOI: 10.1080/19427867.2023.2270238
This paper tackles a complex logistics challenge of disaster management, encompassing warehouse location, pre-disaster inventory planning, routing, and post-disaster relief supply delivery. We establish an iterative process for optimizing relief distribution to shelters. Adaptable warehouse inventory reallocation responds to fluctuating demands, guided by a two-phase mathematical programming approach. In the first phase, a two-stage stochastic programming (TSSP) model determines optimal warehouse and shelter locations and inventory levels. In the subsequent phase, we introduce a mixed-integer programming (MIP) model to minimize the overall delivery time by making routing decisions. To streamline the process, we introduce a novel enumeration algorithm that trims down route options by considering unavailable links, effectively transforming the MIP model into an assignment-based model. This innovation results in a noticeable 74% reduction in solution time. Further efficiency is achieved by developing a branch-and-cut algorithm for swift MIP resolution. A real-world case study confirms the practicality of our approach.
{"title":"A fast-response mathematical programming approach for delivering disaster relief goods: an earthquake case study","authors":"","doi":"10.1080/19427867.2023.2270238","DOIUrl":"10.1080/19427867.2023.2270238","url":null,"abstract":"<div><div>This paper tackles a complex logistics challenge of disaster management, encompassing warehouse location, pre-disaster inventory planning, routing, and post-disaster relief supply delivery. We establish an iterative process for optimizing relief distribution to shelters. Adaptable warehouse inventory reallocation responds to fluctuating demands, guided by a two-phase mathematical programming approach. In the first phase, a two-stage stochastic programming (TSSP) model determines optimal warehouse and shelter locations and inventory levels. In the subsequent phase, we introduce a mixed-integer programming (MIP) model to minimize the overall delivery time by making routing decisions. To streamline the process, we introduce a novel enumeration algorithm that trims down route options by considering unavailable links, effectively transforming the MIP model into an assignment-based model. This innovation results in a noticeable 74% reduction in solution time. Further efficiency is achieved by developing a branch-and-cut algorithm for swift MIP resolution. A real-world case study confirms the practicality of our approach.</div></div>","PeriodicalId":48974,"journal":{"name":"Transportation Letters-The International Journal of Transportation Research","volume":"16 9","pages":"Pages 1091-1114"},"PeriodicalIF":3.3,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135221723","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 : 2024-10-20DOI: 10.1080/19427867.2023.2262201
Speeding has been acknowledged as a critical determinant in increasing the risk of crashes and their resulting injury severities. This paper employs Global Moran’s I coefficient and local Getis – Ord G* indexes to systematically account for the spatial distribution feature of speeding-related crashes, study the global spatial pattern of speeding-related crashes, and identify severe crash cluster districts. The findings demonstrate that severe speeding-related crashes within the state of Pennsylvania have a spatial clustering trend, where four crash datasets are extracted from four hotspot districts. Two log-likelihood ratio (LR) tests were conducted to determine whether speeding-related crashes classified by hotspot districts should be modeled separately. The results suggest that separate modeling is necessary. To capture the unobserved heterogeneity, four correlated random parameter order models with heterogeneity in means are employed to explore the factors contributing to crash severity involving at least one vehicle speeding.
超速已被公认为是增加撞车风险及其造成的严重伤害的关键决定因素。本文采用全局莫兰 I 系数和局部 Getis - Ord G* 指数系统地解释了超速相关碰撞事故的空间分布特征,研究了超速相关碰撞事故的全局空间模式,并确定了严重碰撞事故集群区。研究结果表明,宾夕法尼亚州内与超速相关的严重碰撞事故具有空间集群趋势,其中从四个热点地区提取了四个碰撞事故数据集。为了确定按热点地区分类的超速相关碰撞事故是否应单独建模,进行了两次对数似然比(LR)检验。结果表明,单独建模是必要的。为了捕捉未观察到的异质性,我们采用了四个具有均值异质性的相关随机参数阶次模型,以探讨导致至少有一辆车超速的碰撞严重程度的因素。
{"title":"Investigating the spatial heterogeneity of factors influencing speeding-related crash severities using correlated random parameter order models with heterogeneity-in-means","authors":"","doi":"10.1080/19427867.2023.2262201","DOIUrl":"10.1080/19427867.2023.2262201","url":null,"abstract":"<div><div>Speeding has been acknowledged as a critical determinant in increasing the risk of crashes and their resulting injury severities. This paper employs Global Moran’s I coefficient and local Getis – Ord G* indexes to systematically account for the spatial distribution feature of speeding-related crashes, study the global spatial pattern of speeding-related crashes, and identify severe crash cluster districts. The findings demonstrate that severe speeding-related crashes within the state of Pennsylvania have a spatial clustering trend, where four crash datasets are extracted from four hotspot districts. Two log-likelihood ratio (LR) tests were conducted to determine whether speeding-related crashes classified by hotspot districts should be modeled separately. The results suggest that separate modeling is necessary. To capture the unobserved heterogeneity, four correlated random parameter order models with heterogeneity in means are employed to explore the factors contributing to crash severity involving at least one vehicle speeding.</div></div>","PeriodicalId":48974,"journal":{"name":"Transportation Letters-The International Journal of Transportation Research","volume":"16 9","pages":"Pages 989-1001"},"PeriodicalIF":3.3,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135548622","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 : 2024-10-20DOI: 10.1080/19427867.2023.2275855
The innovations of battery and charging technologies facilitate worldwide bus electrification, which can conserve energy and reduce emissions. The emergence of wireless power transfer technology offers the potential to alleviate bus range anxiety. This paper constructs the layout models of static wireless and centralized charging stations based on the seasonal differences associated with battery capacities and passenger flow, which aims to obtain the bus charging station layout scheme of the lowest cost. Then, the heuristic algorithm is utilized to solve the layout models as the mixed linear programming problem. This study results confirm that the combined layout of static wireless and centralized charging stations is more advantageous than a single model. Finally, the required fixed cost of supporting bus system operation is explored for future capital reserves.
{"title":"Optimizing electric bus charging infrastructure deployment under two charging modes considering seasonal differences","authors":"","doi":"10.1080/19427867.2023.2275855","DOIUrl":"10.1080/19427867.2023.2275855","url":null,"abstract":"<div><div>The innovations of battery and charging technologies facilitate worldwide bus electrification, which can conserve energy and reduce emissions. The emergence of wireless power transfer technology offers the potential to alleviate bus range anxiety. This paper constructs the layout models of static wireless and centralized charging stations based on the seasonal differences associated with battery capacities and passenger flow, which aims to obtain the bus charging station layout scheme of the lowest cost. Then, the heuristic algorithm is utilized to solve the layout models as the mixed linear programming problem. This study results confirm that the combined layout of static wireless and centralized charging stations is more advantageous than a single model. Finally, the required fixed cost of supporting bus system operation is explored for future capital reserves.</div></div>","PeriodicalId":48974,"journal":{"name":"Transportation Letters-The International Journal of Transportation Research","volume":"16 9","pages":"Pages 1126-1139"},"PeriodicalIF":3.3,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135679849","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 : 2024-10-20DOI: 10.1080/19427867.2023.2264048
Pre-peak fare discount policies have become increasingly common in urban transit services and show potential to spread morning peak-hour demand. This study analyzes the departure time choice behavior of interregional bus passengers during morning hours. A mixed logit model is applied to estimate the probability of shifting departure time as a function of fare discount rate, in-vehicle congestion, and other influential factors, accommodating heterogeneity of passenger preferences. The analysis data are obtained through a stated preference survey, which is conducted to passengers who regularly use the interregional bus in Seoul metropolitan area. Choice models are segmented by departure time periods (peak, pre-peak, and post-peak) and occupation types (fixed time worker, flexible time worker, and student). Marginal utilities of the unobserved preferences are estimated to suggest policy implications. From the experimental analysis, target passengers for shifting departure time during peak hour are suggested as flexible time workers and students.
{"title":"Pre-peak fare discount policy for managing morning peak demand of interregional bus travel: a case study in Seoul metropolitan area","authors":"","doi":"10.1080/19427867.2023.2264048","DOIUrl":"10.1080/19427867.2023.2264048","url":null,"abstract":"<div><div>Pre-peak fare discount policies have become increasingly common in urban transit services and show potential to spread morning peak-hour demand. This study analyzes the departure time choice behavior of interregional bus passengers during morning hours. A mixed logit model is applied to estimate the probability of shifting departure time as a function of fare discount rate, in-vehicle congestion, and other influential factors, accommodating heterogeneity of passenger preferences. The analysis data are obtained through a stated preference survey, which is conducted to passengers who regularly use the interregional bus in Seoul metropolitan area. Choice models are segmented by departure time periods (peak, pre-peak, and post-peak) and occupation types (fixed time worker, flexible time worker, and student). Marginal utilities of the unobserved preferences are estimated to suggest policy implications. From the experimental analysis, target passengers for shifting departure time during peak hour are suggested as flexible time workers and students.</div></div>","PeriodicalId":48974,"journal":{"name":"Transportation Letters-The International Journal of Transportation Research","volume":"16 9","pages":"Pages 1059-1068"},"PeriodicalIF":3.3,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135740404","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 : 2024-10-20DOI: 10.1080/19427867.2023.2260972
This study proposes a new method for evaluating the road network vulnerability per unit time: the real-time vulnerability index (RVI). The practical application of the proposed measure in real traffic scenarios is investigated by using empirical data from Shanghai Expressway.To prevent traffic congestion, a critical threshold of RVI (RVI*) is suggested for traffic control. Furthermore, the hysteresis on macroscopic fundamental diagram (MFD) and RVI is further analyzed in the phase transition of traffic states. The results show that RVI can help derive some representative traffic characteristics in real time and dynamically reflect road network vulnerability: when traffic congestion or accidents occur on the road network, the results observed for the traffic hysteresis on MFD and RVI were consistent, and RVI is easier to observe and more suitable for traffic management. The findings prove that RVI is a reliable indicator to help predict the dynamic operation of the road network.
{"title":"Real-time Vulnerability Analysis of Urban Expressway","authors":"","doi":"10.1080/19427867.2023.2260972","DOIUrl":"10.1080/19427867.2023.2260972","url":null,"abstract":"<div><div>This study proposes a new method for evaluating the road network vulnerability per unit time: the real-time vulnerability index (RVI). The practical application of the proposed measure in real traffic scenarios is investigated by using empirical data from Shanghai Expressway.To prevent traffic congestion, a critical threshold of RVI (RVI*) is suggested for traffic control. Furthermore, the hysteresis on macroscopic fundamental diagram (MFD) and RVI is further analyzed in the phase transition of traffic states. The results show that RVI can help derive some representative traffic characteristics in real time and dynamically reflect road network vulnerability: when traffic congestion or accidents occur on the road network, the results observed for the traffic hysteresis on MFD and RVI were consistent, and RVI is easier to observe and more suitable for traffic management. The findings prove that RVI is a reliable indicator to help predict the dynamic operation of the road network.</div></div>","PeriodicalId":48974,"journal":{"name":"Transportation Letters-The International Journal of Transportation Research","volume":"16 9","pages":"Pages 959-977"},"PeriodicalIF":3.3,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136154153","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 : 2024-10-20DOI: 10.1080/19427867.2023.2266189
The objective of the current study is to analyze the time spent on leisure activities in Budapest, considering five influencing factors. Data were collected from Google Popular Time (GPT) via location services using Python, resulting in a dataset of 1336 entries from July 2022. The analysis utilized the Ordered Logit Model (OLM). According to the outcomes, about 17% of visitors allocate significant time to leisure, while half spend relatively less. Leisure time is positively influenced by ratings and location but negatively affected by security levels. This study demonstrates the utility of GPT data for understanding individual behavior, offering valuable insights for decision-makers, tourism managers, and planners. Additionally, it sheds light on leisure-related traffic patterns, aiding in the identification of popular locations and peak time periods for leisure activities. This information can indirectly impact traffic flow in specific areas.
{"title":"Analysis of the factors affecting the time spent on leisure activities by using an ordered logit model","authors":"","doi":"10.1080/19427867.2023.2266189","DOIUrl":"10.1080/19427867.2023.2266189","url":null,"abstract":"<div><div>The objective of the current study is to analyze the time spent on leisure activities in Budapest, considering five influencing factors. Data were collected from Google Popular Time (GPT) via location services using Python, resulting in a dataset of 1336 entries from July 2022. The analysis utilized the Ordered Logit Model (OLM). According to the outcomes, about 17% of visitors allocate significant time to leisure, while half spend relatively less. Leisure time is positively influenced by ratings and location but negatively affected by security levels. This study demonstrates the utility of GPT data for understanding individual behavior, offering valuable insights for decision-makers, tourism managers, and planners. Additionally, it sheds light on leisure-related traffic patterns, aiding in the identification of popular locations and peak time periods for leisure activities. This information can indirectly impact traffic flow in specific areas.</div></div>","PeriodicalId":48974,"journal":{"name":"Transportation Letters-The International Journal of Transportation Research","volume":"16 9","pages":"Pages 1081-1090"},"PeriodicalIF":3.3,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135096370","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 : 2024-10-20DOI: 10.1080/19427867.2023.2261706
The constraints of road network topology and dynamically changing traffic states over time make the task of traffic flow prediction extremely challenging. Most existing methods use CNNs or GCNs to capture spatial correlation. However, convolution operator-based methods are far from optimal in their ability to fuse node features and topology to adequately model spatial correlation. In order to model the spatio-temporal features of traffic flow more effectively, this paper proposes a traffic flow prediction model, the Spatio-Temporal Graph Attention Network (STGAN), which is based on graph attention mechanisms and residually connected gated recurrent units. Specifically, a graph attention mechanism and a random wandering mechanism are used to extract spatial features of the traffic network, and gated recurrent units with residual connections are used to extract temporal features. Experimental results on real-world public transportation datasets show that our approach not only yields state-of-the-art performance, but also exhibits competitive computational efficiency and improves the accuracy of traffic flow prediction.
{"title":"Spatio-temporal graph attention networks for traffic prediction","authors":"","doi":"10.1080/19427867.2023.2261706","DOIUrl":"10.1080/19427867.2023.2261706","url":null,"abstract":"<div><div>The constraints of road network topology and dynamically changing traffic states over time make the task of traffic flow prediction extremely challenging. Most existing methods use CNNs or GCNs to capture spatial correlation. However, convolution operator-based methods are far from optimal in their ability to fuse node features and topology to adequately model spatial correlation. In order to model the spatio-temporal features of traffic flow more effectively, this paper proposes a traffic flow prediction model, the Spatio-Temporal Graph Attention Network (STGAN), which is based on graph attention mechanisms and residually connected gated recurrent units. Specifically, a graph attention mechanism and a random wandering mechanism are used to extract spatial features of the traffic network, and gated recurrent units with residual connections are used to extract temporal features. Experimental results on real-world public transportation datasets show that our approach not only yields state-of-the-art performance, but also exhibits competitive computational efficiency and improves the accuracy of traffic flow prediction.</div></div>","PeriodicalId":48974,"journal":{"name":"Transportation Letters-The International Journal of Transportation Research","volume":"16 9","pages":"Pages 978-988"},"PeriodicalIF":3.3,"publicationDate":"2024-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136279553","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}