Pub Date : 2026-01-29DOI: 10.1016/j.trd.2026.105226
Jialu Yang, Danesh Hosseinpanahi, Bo Zou, Jane Lin
Hydrogen fuel cell trucks (HFCTs) are a promising alternative to diesel trucks (DTs) for decarbonizing long-haul freight while retaining comparable operational performance. This study investigates a large scale HRS siting problem with the aim of minimizing the capital investment in HRS construction and hydrogen delivery cost. A hybrid delivery method is considered: hydrogen is delivered from production hubs by the existing natural gas pipeline network to storage reservoirs, then by truck to HRSs. Assuming 10% of the 2050 Freight Analysis Framework Version 5 (FAF5) truck flow as HFCTs on the U.S. continental interstate highway network, results show that pipeline-based delivery offers major economic advantages over truck-based delivery. Findings highlight the strategic value of leveraging existing pipeline infrastructure and suggest that targeted policy support for hydrogen delivery and refueling infrastructure is essential for facilitating the cost-effective adoption of HFCTs in long-haul freight.
{"title":"Powering long haul freight: hydrogen refueling station siting using pipeline infrastructure","authors":"Jialu Yang, Danesh Hosseinpanahi, Bo Zou, Jane Lin","doi":"10.1016/j.trd.2026.105226","DOIUrl":"10.1016/j.trd.2026.105226","url":null,"abstract":"<div><div>Hydrogen fuel cell trucks (HFCTs) are a promising alternative to diesel trucks (DTs) for decarbonizing long-haul freight while retaining comparable operational performance. This study investigates a large scale HRS siting problem with the aim of minimizing the capital investment in HRS construction and hydrogen delivery cost. A hybrid delivery method is considered: hydrogen is delivered from production hubs by the existing natural gas pipeline network to storage reservoirs, then by truck to HRSs. Assuming 10% of the 2050 Freight Analysis Framework Version 5 (FAF5) truck flow as HFCTs on the U.S. continental interstate highway network, results show that pipeline-based delivery offers major economic advantages over truck-based delivery. Findings highlight the strategic value of leveraging existing pipeline infrastructure and suggest that targeted policy support for hydrogen delivery and refueling infrastructure is essential for facilitating the cost-effective adoption of HFCTs in long-haul freight.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"154 ","pages":"Article 105226"},"PeriodicalIF":7.7,"publicationDate":"2026-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146070871","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-01-28DOI: 10.1016/j.trd.2025.105168
Jiayou Lei , Mingwei He , Min He , Zhuangbin Shi , Yueren He , Yang Liu , Huimin Qian
Understanding intermodal transit trip patterns is essential for advancing bus-metro integration. However, most existing studies focus on aggregate transit ridership, while empirical evidence on the characteristics of intermodal transit trips and the effects of the built environment (BE) on them remains limited. To address this gap, this study proposes a probabilistic topic modeling-artificial neural network (ANN) framework to identify representative intermodal transit trip patterns and examine their interactions with BE, using one week of large-scale smart card data from Beijing, China. A multidimensional latent Dirichlet allocation (LDA) model extracts grid-level patterns, revealing five spatiotemporal profiles with probabilistic interpretations. A multi-input, multi-output ANN, interpreted via SHapley Additive exPlanations (SHAP), uncovers complex nonlinear relationships and threshold effects between BE and trip pattern distributions. Results demonstrate that BE influences trip pattern distributions in ways distinct from its effects on overall ridership, providing new insights for coordinated bus-metro development and informing targeted integration strategies.
{"title":"Quantifying built environment effects on intermodal transit trip patterns at grid level","authors":"Jiayou Lei , Mingwei He , Min He , Zhuangbin Shi , Yueren He , Yang Liu , Huimin Qian","doi":"10.1016/j.trd.2025.105168","DOIUrl":"10.1016/j.trd.2025.105168","url":null,"abstract":"<div><div>Understanding intermodal transit trip patterns is essential for advancing bus-metro integration. However, most existing studies focus on aggregate transit ridership, while empirical evidence on the characteristics of intermodal transit trips and the effects of the built environment (BE) on them remains limited. To address this gap, this study proposes a probabilistic topic modeling-artificial neural network (ANN) framework to identify representative intermodal transit trip patterns and examine their interactions with BE, using one week of large-scale smart card data from Beijing, China. A multidimensional latent Dirichlet allocation (LDA) model extracts grid-level patterns, revealing five spatiotemporal profiles with probabilistic interpretations. A multi-input, multi-output ANN, interpreted via SHapley Additive exPlanations (SHAP), uncovers complex nonlinear relationships and threshold effects between BE and trip pattern distributions. Results demonstrate that BE influences trip pattern distributions in ways distinct from its effects on overall ridership, providing new insights for coordinated bus-metro development and informing targeted integration strategies.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"153 ","pages":"Article 105168"},"PeriodicalIF":7.7,"publicationDate":"2026-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146071722","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-01-28DOI: 10.1016/j.trd.2026.105246
Liyang Huang , Liyuan Zhao , Zhi-Chun Li
Urban logistics have shifted from a “goods-to-store” to a “goods-to-residents” paradigm, resulting in the proliferation of small logistics facilities such as micro-hubs and pick-up points in urban areas. As these facilities are increasingly embedded within urban functional spaces, understanding the driving factors of their spatial layout has become ever more crucial. This study examines the spatial layout characteristics and influencing factors of small logistics facilities in the Wuhan Metropolitan Development Area between 2010 and 2018. A Gradient Boosting Decision Tree (GBDT) model is employed to analyze the nonlinear relationships and threshold effects of urban spatial variables on facility layout. The results indicate that land price exerts the strongest influence on spatial distribution, followed by transportation accessibility and demand-related variables in 2010 and 2018, respectively. Based on these findings, a decision-support tool is developed to optimize facility placement, offering policy-relevant insights for sustainable urban logistics planning from a supply–demand perspective.
{"title":"Effects of functional spaces on small logistics facilities: machine learning-based decision-support tool","authors":"Liyang Huang , Liyuan Zhao , Zhi-Chun Li","doi":"10.1016/j.trd.2026.105246","DOIUrl":"10.1016/j.trd.2026.105246","url":null,"abstract":"<div><div>Urban logistics have shifted from a “goods-to-store” to a “goods-to-residents” paradigm, resulting in the proliferation of small logistics facilities such as micro-hubs and pick-up points in urban areas. As these facilities are increasingly embedded within urban functional spaces, understanding the driving factors of their spatial layout has become ever more crucial. This study examines the spatial layout characteristics and influencing factors of small logistics facilities in the Wuhan Metropolitan Development Area between 2010 and 2018. A Gradient Boosting Decision Tree (GBDT) model is employed to analyze the nonlinear relationships and threshold effects of urban spatial variables on facility layout. The results indicate that land price exerts the strongest influence on spatial distribution, followed by transportation accessibility and demand-related variables in 2010 and 2018, respectively. Based on these findings, a decision-support tool is developed to optimize facility placement, offering policy-relevant insights for sustainable urban logistics planning from a supply–demand perspective.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"153 ","pages":"Article 105246"},"PeriodicalIF":7.7,"publicationDate":"2026-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146072514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-01-28DOI: 10.1016/j.trd.2026.105242
Corrado Maria Caminiti , Davide Fratelli , Aleksandar Dimovski , Antonio Piersanti , Mattia Ricci , Biagio Di Pietra , Marco Merlo
Achieving decarbonization targets requires major reductions in fossil-fuel use. Electrifying light-duty vehicles offers substantial environmental benefits, yet quantifying these gains is challenging due to heterogeneous traffic patterns and pollution levels. This study introduces a GIS-based method that maps pollutant emissions and traffic volumes onto the road network to produce municipal-level mobility-impact indexes. Daily travel routines derived from mobility surveys are combined with real pollution maps to compute segment-level emissions, which are then aggregated for spatial comparison and policy prioritisation. Applied to Lombardy, Italy, the method shows that although Milan and other dense cities benefit markedly, numerous peri-urban and semi-rural municipalities also show high potential due to large intermunicipal commuting flows. Overall, 13% of municipalities are classified as highest-priority areas, accounting for nearly a quarter of regional traffic-related emissions. The results reveal non-linear urbanisation-benefit relationships and underscore the need to include less-urbanised territories in equitable and effective electrification strategies.
{"title":"Prioritizing light-vehicle electrification cities through real-world pollution-based GIS assessment","authors":"Corrado Maria Caminiti , Davide Fratelli , Aleksandar Dimovski , Antonio Piersanti , Mattia Ricci , Biagio Di Pietra , Marco Merlo","doi":"10.1016/j.trd.2026.105242","DOIUrl":"10.1016/j.trd.2026.105242","url":null,"abstract":"<div><div>Achieving decarbonization targets requires major reductions in fossil-fuel use. Electrifying light-duty vehicles offers substantial environmental benefits, yet quantifying these gains is challenging due to heterogeneous traffic patterns and pollution levels. This study introduces a GIS-based method that maps pollutant emissions and traffic volumes onto the road network to produce municipal-level mobility-impact indexes. Daily travel routines derived from mobility surveys are combined with real pollution maps to compute segment-level emissions, which are then aggregated for spatial comparison and policy prioritisation. Applied to Lombardy, Italy, the method shows that although Milan and other dense cities benefit markedly, numerous peri-urban and semi-rural municipalities also show high potential due to large intermunicipal commuting flows. Overall, 13% of municipalities are classified as highest-priority areas, accounting for nearly a quarter of regional traffic-related emissions. The results reveal non-linear urbanisation-benefit relationships and underscore the need to include less-urbanised territories in equitable and effective electrification strategies.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"153 ","pages":"Article 105242"},"PeriodicalIF":7.7,"publicationDate":"2026-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146071723","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-01-28DOI: 10.1016/j.trd.2026.105244
Robert B. Noland
{"title":"The impact of research on transportation and the environment","authors":"Robert B. Noland","doi":"10.1016/j.trd.2026.105244","DOIUrl":"10.1016/j.trd.2026.105244","url":null,"abstract":"","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"153 ","pages":"Article 105244"},"PeriodicalIF":7.7,"publicationDate":"2026-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146072513","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-01-27DOI: 10.1016/j.trd.2026.105233
Bingkun Chen , Zhuo Chen , Xiaoyue Cathy Liu , Ran Wei , Arman Malekloo
Paratransit, a demand-responsive transit mode serving passengers with mobility challenges, is increasingly electrified to enhance urban transportation sustainability. However, high investments required for dedicated charging infrastructure and the scarcity of public charging resources remain significant hurdles to large-scale deployment. This study investigates a shared charging scheme that integrates paratransit electric vehicles (EVs) into existing electric bus (EB) charging networks. A fuzzy multi-objective optimization framework is proposed to identify optimal charging co-hub locations and EV assignments by balancing supply–demand dynamics. The framework incorporates two-step floating catchment area (2SFCA) and inverted 2SFCA (i2SFCA) methods to formulate objectives and constraints for EB and paratransit systems, respectively. Through fuzzy programming, trade-offs among supply–demand dynamics are resolved, yielding efficient shared-charging plans. The framework is validated with Utah Transit Authority data, demonstrating improved charging accessibility and operational efficiency while offering actionable insights for transit agencies in planning shared charging schemes among various public transport modes.
{"title":"Co-hub charging planning for electric bus and paratransit using fuzzy multi-objective optimization","authors":"Bingkun Chen , Zhuo Chen , Xiaoyue Cathy Liu , Ran Wei , Arman Malekloo","doi":"10.1016/j.trd.2026.105233","DOIUrl":"10.1016/j.trd.2026.105233","url":null,"abstract":"<div><div>Paratransit, a demand-responsive transit mode serving passengers with mobility challenges, is increasingly electrified to enhance urban transportation sustainability. However, high investments required for dedicated charging infrastructure and the scarcity of public charging resources remain significant hurdles to large-scale deployment. This study investigates a shared charging scheme that integrates paratransit electric vehicles (EVs) into existing electric bus (EB) charging networks. A fuzzy multi-objective optimization framework is proposed to identify optimal charging co-hub locations and EV assignments by balancing supply–demand dynamics. The framework incorporates two-step floating catchment area (2SFCA) and inverted 2SFCA (i2SFCA) methods to formulate objectives and constraints for EB and paratransit systems, respectively. Through fuzzy programming, trade-offs among supply–demand dynamics are resolved, yielding efficient shared-charging plans. The framework is validated with Utah Transit Authority data, demonstrating improved charging accessibility and operational efficiency while offering actionable insights for transit agencies in planning shared charging schemes among various public transport modes.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"153 ","pages":"Article 105233"},"PeriodicalIF":7.7,"publicationDate":"2026-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146072519","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-01-24DOI: 10.1016/j.trd.2026.105236
Quanxu Zhou, Mengying Huang, Haoyu Tang, Jiapeng Cheng, Ji Wu
As urban transportation electrification advances, range anxiety has become a key concern for electric vehicles (EVs). Battery swapping stations (BSS) offer a fast and efficient solution, yet traditional scheduling methods struggle to balance profitability with operational risk in dynamic environments. To address this, we propose a data-driven optimization method. First, historical data is used to derive optimal operational decisions, and an EV battery swapping demand forecasting model is built. Based on the forecasted demand and historical strategies, a long short-term memory network predicts the BSS’s overall charging and discharging power. A double deep Q-network is then employed to allocate this power to individual batteries, ensuring timely swaps. Validation using real operational data from Chengdu, China, shows the proposed method effectively meets battery swapping demand, enhances scheduling efficiency and station profitability, and reduces peak loads and power fluctuations, demonstrating the potential for practical application in managing smart EV infrastructure.
{"title":"Profit-aware battery swapping station energy scheduling via hybrid hierarchical deep reinforcement learning","authors":"Quanxu Zhou, Mengying Huang, Haoyu Tang, Jiapeng Cheng, Ji Wu","doi":"10.1016/j.trd.2026.105236","DOIUrl":"10.1016/j.trd.2026.105236","url":null,"abstract":"<div><div>As urban transportation electrification advances, range anxiety has become a key concern for electric vehicles (EVs). Battery swapping stations (BSS) offer a fast and efficient solution, yet traditional scheduling methods struggle to balance profitability with operational risk in dynamic environments. To address this, we propose a data-driven optimization method. First, historical data is used to derive optimal operational decisions, and an EV battery swapping demand forecasting model is built. Based on the forecasted demand and historical strategies, a long short-term memory network predicts the BSS’s overall charging and discharging power. A double deep Q-network is then employed to allocate this power to individual batteries, ensuring timely swaps. Validation using real operational data from Chengdu, China, shows the proposed method effectively meets battery swapping demand, enhances scheduling efficiency and station profitability, and reduces peak loads and power fluctuations, demonstrating the potential for practical application in managing smart EV infrastructure.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"153 ","pages":"Article 105236"},"PeriodicalIF":7.7,"publicationDate":"2026-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146038813","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Road network expansion is a key factor driving fragmentation of wildlife habitats and threatening the biodiversity of national parks. Ecological corridors identification and wildlife crossing site selection based on MaxEnt and MCR model are an innovative pathway to coordinate the conflict. Taking Hainan Tropical Rainforest National Park and its circle road as a case, 145 ecological habitat sources of terrestrial, arboreal and amphibious reptile animals were identified, 298 ecological corridors were simulated using Linkage Mapper, and 585 conflict points between road and ecological corridor are identified with verification. Based on distance thresholds and species-specific behaviors, 274 wildlife crossing sites of terrestrial, arboreal and amphibian crossing were laid out along road with 1.7 km spacing, and 4 kind of wildlife crossing were designed according to animal habits. This study provides a theoretical framework for the combination of transportation network with ecological protection with wildlife crossings to improve the biodiversity of nature reserves.
{"title":"Road and eco-corridor conflict mitigation through multi-species wildlife crossings around national park","authors":"Yuting Peng , Hongwei Zhang , Gaoru Zhu , Xing Yang , Xueyan Zhao , Huaping Liang","doi":"10.1016/j.trd.2026.105235","DOIUrl":"10.1016/j.trd.2026.105235","url":null,"abstract":"<div><div>Road network expansion is a key factor driving fragmentation of wildlife habitats and threatening the biodiversity of national parks. Ecological corridors identification and wildlife crossing site selection based on MaxEnt and MCR model are an innovative pathway to coordinate the conflict. Taking Hainan Tropical Rainforest National Park and its circle road as a case, 145 ecological habitat sources of terrestrial, arboreal and amphibious reptile animals were identified, 298 ecological corridors were simulated using Linkage Mapper, and 585 conflict points between road and ecological corridor are identified with verification. Based on distance thresholds and species-specific behaviors, 274 wildlife crossing sites of terrestrial, arboreal and amphibian crossing were laid out along road with 1.7 km spacing, and 4 kind of wildlife crossing were designed according to animal habits. This study provides a theoretical framework for the combination of transportation network with ecological protection with wildlife crossings to improve the biodiversity of nature reserves.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"153 ","pages":"Article 105235"},"PeriodicalIF":7.7,"publicationDate":"2026-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146038814","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-01-24DOI: 10.1016/j.trd.2026.105232
Ruolin Yao , Shuai Ling , Qing Wang , Shiquan Zhong
To alleviate range anxiety among electric vehicle (EV) users, some EV manufacturers have begun operating charging stations in public areas, offering charging services to vehicle owners. Based on data from the Chinese EV market, we investigate how the strategy of operating charging stations influences demand for manufacturers’ own-brand EVs and examine its interaction with EV range, as well as the moderating effect of public charging infrastructure. Specifically, we integrate the strategy of operating charging stations as a model-specific charging service attribute, measure it using the scale of the charging stations, and incorporate it into a structural model. Our results indicate that this strategy enhances consumers’ willingness to purchase, although it increases their price sensitivity. Moreover, this willingness to purchase increases with EV range but decreases as public charging infrastructure improves. These findings offer valuable insights for the collaborative development of the EV and charging station industries.
{"title":"Electric vehicle manufacturers’ operation of charging stations: Higher range, greater benefits","authors":"Ruolin Yao , Shuai Ling , Qing Wang , Shiquan Zhong","doi":"10.1016/j.trd.2026.105232","DOIUrl":"10.1016/j.trd.2026.105232","url":null,"abstract":"<div><div>To alleviate range anxiety among electric vehicle (EV) users, some EV manufacturers have begun operating charging stations in public areas, offering charging services to vehicle owners. Based on data from the Chinese EV market, we investigate how the strategy of operating charging stations influences demand for manufacturers’ own-brand EVs and examine its interaction with EV range, as well as the moderating effect of public charging infrastructure. Specifically, we integrate the strategy of operating charging stations as a model-specific charging service attribute, measure it using the scale of the charging stations, and incorporate it into a structural model. Our results indicate that this strategy enhances consumers’ willingness to purchase, although it increases their price sensitivity. Moreover, this willingness to purchase increases with EV range but decreases as public charging infrastructure improves. These findings offer valuable insights for the collaborative development of the EV and charging station industries.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"153 ","pages":"Article 105232"},"PeriodicalIF":7.7,"publicationDate":"2026-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146038812","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-01-24DOI: 10.1016/j.trd.2026.105241
Zirun Wang, Huasa Zhu, Chuanqi Ma, Zhaojie Lu, Siqing Long, Yunjie Zhang, Ming Cai
Vehicles with illegally modified exhaust systems are a significant source of urban noise and air pollution. This study employs a Sound Event Detection (SED) framework for their automated identification. To address the lack of public data, we generated a dataset with Scaper library, embedding foreground events into authentic background recordings. We employed an SED Transformer and compared six backbones, with the Multi-Scale Residual Network (MSResNet) achieving the best performance. The model achieved an event-based F1-score of 0.7277 on the validation set and 0.6629 in the field test. A Shapley Additive Explanations (SHAP) analysis confirmed the model’s focus on meaningful acoustic features like acceleration harmonics and afterfire transients, while also revealing a temporal bias from the synthetic data. Analysis of computational performance and an event-covered localization error metric validated deployment feasibility. This work presents an end-to-end SED framework for automated urban noise enforcement, supporting data-driven policy beyond conventional classification.
{"title":"Sound event detection for modified-exhaust vehicles in urban environment","authors":"Zirun Wang, Huasa Zhu, Chuanqi Ma, Zhaojie Lu, Siqing Long, Yunjie Zhang, Ming Cai","doi":"10.1016/j.trd.2026.105241","DOIUrl":"10.1016/j.trd.2026.105241","url":null,"abstract":"<div><div>Vehicles with illegally modified exhaust systems are a significant source of urban noise and air pollution. This study employs a Sound Event Detection (SED) framework for their automated identification. To address the lack of public data, we generated a dataset with Scaper library, embedding foreground events into authentic background recordings. We employed an SED Transformer and compared six backbones, with the Multi-Scale Residual Network (MSResNet) achieving the best performance. The model achieved an event-based F1-score of 0.7277 on the validation set and 0.6629 in the field test. A Shapley Additive Explanations (SHAP) analysis confirmed the model’s focus on meaningful acoustic features like acceleration harmonics and afterfire transients, while also revealing a temporal bias from the synthetic data. Analysis of computational performance and an event-covered localization error metric validated deployment feasibility. This work presents an end-to-end SED framework for automated urban noise enforcement, supporting data-driven policy beyond conventional classification.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"153 ","pages":"Article 105241"},"PeriodicalIF":7.7,"publicationDate":"2026-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146048442","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号