Xiaobo Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Xiaobo Liu has a total of 155 co-authors. Here are the top co-authors by number of shared publications:
- Fangfang Zheng
School of Transportation and Logistics, Southwest Jiaotong University, Chengdu, China
15 shared publications - Peng Cao
Southwest Jiaotong University
10 shared publications - Zhandong Xu
Southwest Jiaotong University
10 shared publications - Mi Gan
School of Transportation and Logistics, Southwest Jiaotong University, Chengdu, China
5 shared publications Peng Cao
5 shared publicationsRuijie Li
5 shared publications
Latest Publications
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 27 (11) , 14976-14988
Efficiency Versus Revenue: Bi-Objective Reinforcement Learning for Coordinated Ramp Metering
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 27 (11) , 14976-14988
Efficiency Versus Revenue: Bi-Objective Reinforcement Learning for Coordinated Ramp Metering
- Open Access
- 16 October 2026
- Published by Springer Nature in Railway Engineering Science
- p. 1-25
An exact approach to the flat shunting problem with a focus on car movement cost optimization
- 16 October 2026
- Published by Springer Nature in Railway Engineering Science
- p. 1-25
Open AccessAn exact approach to the flat shunting problem with a focus on car movement cost optimization
- 19 September 2026
- Published by Elsevier in Safety Science
- Vol. 205
Large-scale experimental study of stair evacuation speeds for heterogeneous individuals in high-rise buildings
- 19 September 2026
- Published by Elsevier in Safety Science
- Vol. 205
Large-scale experimental study of stair evacuation speeds for heterogeneous individuals in high-rise buildings
- 18 September 2026
- Published by Elsevier in Accident Analysis & Prevention
- Vol. 238, 108774
On the quantification of the spatiotemporal impact of a single discretionary lane change on upstream traffic safety and efficiency
- 18 September 2026
- Published by Elsevier in Accident Analysis & Prevention
- Vol. 238, 108774
On the quantification of the spatiotemporal impact of a single discretionary lane change on upstream traffic safety and efficiency
- 18 August 2026
- Published by Elsevier in Transportation Research Part C: Emerging Technologies
- Vol. 193
A hybrid optimization framework for cooperative CAV merging based on traveling salesman problem
- 18 August 2026
- Published by Elsevier in Transportation Research Part C: Emerging Technologies
- Vol. 193
A hybrid optimization framework for cooperative CAV merging based on traveling salesman problem
- Open Access
- 1 August 2026
- Published by Elsevier in SSRN Electronic Journal
During-disaster restoration scheduling for transportation networks under uncertain rainstorms
Prolonged rainstorms pose substantial threats to transportation systems and expose the limitations of conventional post-disaster restoration, in which recovery actions are initiated only after the disaster has ended. This study proposes a during-disaster restoration (DDR) scheduling framework that initiates restoration while rainstorm-induced disruptions are still evolving. The framework combines a distributionally robust characterization of rainstorm conditions with restoration scheduling and demand loss evaluation. A case study on a real-world road network in Chengdu, China, evaluates the proposed framework. First, DDR consistently reduces travel demand loss and accelerates network recovery compared with post-disaster restoration. Second, rainstorm uncertainty can substantially affect restoration performance, resulting in considerable deviations from deterministic scenarios. Third, although rainstorm scale determines the overall magnitude of degradation, timely intervention remains the most effective operational strategy for mitigating disaster impacts. These findings provide practical support for transportation agencies in developing adaptive and resilient restoration strategies for extreme weather events.
- 1 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessDuring-disaster restoration scheduling for transportation networks under uncertain rainstorms
Prolonged rainstorms pose substantial threats to transportation systems and expose the limitations of conventional post-disaster restoration, in which recovery actions are initiated only after the disaster has ended. This study proposes a during-disaster restoration (DDR) scheduling framework that initiates restoration while rainstorm-induced disruptions are still evolving. The framework combines a distributionally robust characterization of rainstorm conditions with restoration scheduling and demand loss evaluation. A case study on a real-world road network in Chengdu, China, evaluates the proposed framework. First, DDR consistently reduces travel demand loss and accelerates network recovery compared with post-disaster restoration. Second, rainstorm uncertainty can substantially affect restoration performance, resulting in considerable deviations from deterministic scenarios. Third, although rainstorm scale determines the overall magnitude of degradation, timely intervention remains the most effective operational strategy for mitigating disaster impacts. These findings provide practical support for transportation agencies in developing adaptive and resilient restoration strategies for extreme weather events.
- 3 June 2026
- Published by IOP Publishing in Measurement Science and Technology
- Vol. 37 (28) , 285107
A Roadside LiDAR Placement Optimization Method for Maximizing Vehicle Detection Based on Point Cloud Spatial Distribution at Urban Intersections
Optimizing the placement of roadside LiDAR sensors is critical for achieving reliable vehicle perception in roadside sensing systems.Existing placement optimization methods primarily aim to maximize road-level point cloud coverage. However, such strategies do not necessarily ensure high-quality coverage of vehicle surfaces, which is essential for accurate downstream perception tasks such as detection and tracking. To address this limitation, we propose a LiDAR placement optimization framework that explicitly enhances the spatial distribution of point clouds on vehicle surfaces. First, we introduce a novel metric, termed point cloud coverage (PCC), to quantitatively evaluate the spatial distribution quality of LiDAR points on vehicle surfaces from a geometric perspective. The PCC metric analyzes the spatial relationship between LiDAR beams and vehicle surfaces in a parametric manner and computes surface coverage using an adaptive nearest-neighbor algorithm, enabling accurate and efficient evaluation of sensing performance. Furthermore, we formulate a PCC-guided LiDAR placement optimization model that improves vehicle surface point cloud coverage under practical deployment constraints. Experimental results demonstrate that the proposed analytical metric maintains strong consistency with evaluations based on perception algorithms while significantly reducing computational cost. Experiments conducted at a complex urban intersection show that the optimized multi-LiDAR configurations improve vehicle detection recall by 5.5% to 10.1% compared with baseline placements. Overall, this work reframes roadside LiDAR placement as a point cloud distribution optimization problem and provides a principled framework for improving the spatial quality of acquired data, thereby enhancing the reliability of downstream perception tasks.
- 3 June 2026
- Published by IOP Publishing in Measurement Science and Technology
- Vol. 37 (28) , 285107
A Roadside LiDAR Placement Optimization Method for Maximizing Vehicle Detection Based on Point Cloud Spatial Distribution at Urban Intersections
Optimizing the placement of roadside LiDAR sensors is critical for achieving reliable vehicle perception in roadside sensing systems.Existing placement optimization methods primarily aim to maximize road-level point cloud coverage. However, such strategies do not necessarily ensure high-quality coverage of vehicle surfaces, which is essential for accurate downstream perception tasks such as detection and tracking. To address this limitation, we propose a LiDAR placement optimization framework that explicitly enhances the spatial distribution of point clouds on vehicle surfaces. First, we introduce a novel metric, termed point cloud coverage (PCC), to quantitatively evaluate the spatial distribution quality of LiDAR points on vehicle surfaces from a geometric perspective. The PCC metric analyzes the spatial relationship between LiDAR beams and vehicle surfaces in a parametric manner and computes surface coverage using an adaptive nearest-neighbor algorithm, enabling accurate and efficient evaluation of sensing performance. Furthermore, we formulate a PCC-guided LiDAR placement optimization model that improves vehicle surface point cloud coverage under practical deployment constraints. Experimental results demonstrate that the proposed analytical metric maintains strong consistency with evaluations based on perception algorithms while significantly reducing computational cost. Experiments conducted at a complex urban intersection show that the optimized multi-LiDAR configurations improve vehicle detection recall by 5.5% to 10.1% compared with baseline placements. Overall, this work reframes roadside LiDAR placement as a point cloud distribution optimization problem and provides a principled framework for improving the spatial quality of acquired data, thereby enhancing the reliability of downstream perception tasks.