Ao Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ao Li has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Xin Gao
School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081, China
4 shared publications - Xueyuan Li
School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081, China
4 shared publications - Zhaoyang Ma
Beijing Jiaotong University
3 shared publications Hao Liu
School of Mechanical Engineering, Beijing Institute of Technology, Beijing, China
2 shared publications- Zirui Li
Nanyang Technological University
2 shared publications Changjian Zhao
School of Mechanical Engineering, Beijing Institute of Technology, Beijing, China
1 shared publication
Latest Publications
- 19 May 2026
- Published by Elsevier in Neural Networks
- Vol. 203, 109148
Omni-domain energy-efficient decision-making for large-scale heterogeneous platoons with dual-level graph reinforcement learning
- 19 May 2026
- Published by Elsevier in Neural Networks
- Vol. 203, 109148
Omni-domain energy-efficient decision-making for large-scale heterogeneous platoons with dual-level graph reinforcement learning
- Open Access
- 5 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (20) , 43403-43421
Spatiotemporal Coordinated Decision-Making in Heterogeneous Platooning via Interflow Hub-and-Spoke Graph Reinforcement Learning
In mixed vehicular platooning, heterogeneity from varying vehicle types and control systems leads to unpredictable traffic behaviors and frequent inconsistencies. These variances worsen the spatio-temporal discord within platoons, leading to persistent virtual bottlenecks that substantially impede traffic flow and increase energy consumption. To address these challenges, a centralized decision-making framework utilizing interflow hub-and-spoke graph reinforcement learning (InterHub-GRL) is implemented. This strategy enhances the precision of spatio-temporal interaction models among heterogeneous elements within the platooning, promoting collaborative decision-making, ensuring energy efficiency, and alleviating congestion. The proposed framework captures dynamic interactions within and between platoons in non-Euclidean spaces. Integrating spatio-temporal weighted graphs with a multi-head attention mechanism significantly improves the model’s ability to process both local and global data. Comparative algorithm experiments, generalizability assessments, and permeability ablation studies using the I-24 dataset demonstrate that this strategy achieves a 10% increase in throughput and a 9% reduction in energy consumption compared to baseline metrics. Notably, increasing the penetration rate of connected and autonomous vehicles (CAVs) markedly enhances traffic throughput while managing energy consumption.
- 5 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (20) , 43403-43421
Open AccessSpatiotemporal Coordinated Decision-Making in Heterogeneous Platooning via Interflow Hub-and-Spoke Graph Reinforcement Learning
In mixed vehicular platooning, heterogeneity from varying vehicle types and control systems leads to unpredictable traffic behaviors and frequent inconsistencies. These variances worsen the spatio-temporal discord within platoons, leading to persistent virtual bottlenecks that substantially impede traffic flow and increase energy consumption. To address these challenges, a centralized decision-making framework utilizing interflow hub-and-spoke graph reinforcement learning (InterHub-GRL) is implemented. This strategy enhances the precision of spatio-temporal interaction models among heterogeneous elements within the platooning, promoting collaborative decision-making, ensuring energy efficiency, and alleviating congestion. The proposed framework captures dynamic interactions within and between platoons in non-Euclidean spaces. Integrating spatio-temporal weighted graphs with a multi-head attention mechanism significantly improves the model’s ability to process both local and global data. Comparative algorithm experiments, generalizability assessments, and permeability ablation studies using the I-24 dataset demonstrate that this strategy achieves a 10% increase in throughput and a 9% reduction in energy consumption compared to baseline metrics. Notably, increasing the penetration rate of connected and autonomous vehicles (CAVs) markedly enhances traffic throughput while managing energy consumption.
- Open Access
Torque Differential-Based Dynamic Modeling, Validation, and Steering Characteristics Analysis of Multi-Axial Skid-Steered Wheeled Vehicle
Skid-steered technology has been widely applied to wheeled vehicles due to its advanced pivot steering capabilities and adaptable power transmission systems. To better understand the steering characteristics of skid-steered wheeled vehicles and derive general conclusions, a torque differential-based (TD-based) two-degree-of-freedom (2-DOF) dynamic model was developed. This model is grounded in the vehicle’s steering mechanism and the single-wheel dynamics model, and a steering radius model based on TD input was established using the concept of the instantaneous center of rotation (ICR). Additionally, an expression for the stability factor was derived, and both the steady-state and transient steering characteristics were analyzed. Finally, the real vehicle tests demonstrated that the TD-based dynamic model responds quickly, maintains high precision, and remains stable across different steering input frequencies. Compared with the SD-based dynamic model, the TD-based dynamic model has a 2.553% higher calculation accuracy for the steering radius, a 6.251% higher comprehensive response accuracy for the steering input, and a response speed advantage of 1.035%.
Open AccessTorque Differential-Based Dynamic Modeling, Validation, and Steering Characteristics Analysis of Multi-Axial Skid-Steered Wheeled Vehicle
Skid-steered technology has been widely applied to wheeled vehicles due to its advanced pivot steering capabilities and adaptable power transmission systems. To better understand the steering characteristics of skid-steered wheeled vehicles and derive general conclusions, a torque differential-based (TD-based) two-degree-of-freedom (2-DOF) dynamic model was developed. This model is grounded in the vehicle’s steering mechanism and the single-wheel dynamics model, and a steering radius model based on TD input was established using the concept of the instantaneous center of rotation (ICR). Additionally, an expression for the stability factor was derived, and both the steady-state and transient steering characteristics were analyzed. Finally, the real vehicle tests demonstrated that the TD-based dynamic model responds quickly, maintains high precision, and remains stable across different steering input frequencies. Compared with the SD-based dynamic model, the TD-based dynamic model has a 2.553% higher calculation accuracy for the steering radius, a 6.251% higher comprehensive response accuracy for the steering input, and a response speed advantage of 1.035%.
- Open Access
A Nested Graph Reinforcement Learning-based Decision-making Strategy for Eco-platooning
Platooning technology is renowned for its precise vehicle control, traffic flow optimization, and energy efficiency enhancement. However, in large-scale mixed platoons, vehicle heterogeneity and unpredictable traffic conditions lead to virtual bottlenecks. These bottlenecks result in reduced traffic throughput and increased energy consumption within the platoon. To address these challenges, we introduce a decision-making strategy based on nested graph reinforcement learning. This strategy improves collaborative decision-making, ensuring energy efficiency and alleviating congestion. We propose a theory of nested traffic graph representation that maps dynamic interactions between vehicles and platoons in non-Euclidean spaces. By incorporating spatio-temporal weighted graph into a multi-head attention mechanism, we further enhance the model's capacity to process both local and global data. Additionally, we have developed a nested graph reinforcement learning framework to enhance the self-iterative learning capabilities of platooning. Using the I-24 dataset, we designed and conducted comparative algorithm experiments, generalizability testing, and permeability ablation experiments, thereby validating the proposed strategy's effectiveness. Compared to the baseline, our strategy increases throughput by 10% and decreases energy use by 9%. Specifically, increasing the penetration rate of CAVs significantly enhances traffic throughput, though it also increases energy consumption.
Open AccessA Nested Graph Reinforcement Learning-based Decision-making Strategy for Eco-platooning
Platooning technology is renowned for its precise vehicle control, traffic flow optimization, and energy efficiency enhancement. However, in large-scale mixed platoons, vehicle heterogeneity and unpredictable traffic conditions lead to virtual bottlenecks. These bottlenecks result in reduced traffic throughput and increased energy consumption within the platoon. To address these challenges, we introduce a decision-making strategy based on nested graph reinforcement learning. This strategy improves collaborative decision-making, ensuring energy efficiency and alleviating congestion. We propose a theory of nested traffic graph representation that maps dynamic interactions between vehicles and platoons in non-Euclidean spaces. By incorporating spatio-temporal weighted graph into a multi-head attention mechanism, we further enhance the model's capacity to process both local and global data. Additionally, we have developed a nested graph reinforcement learning framework to enhance the self-iterative learning capabilities of platooning. Using the I-24 dataset, we designed and conducted comparative algorithm experiments, generalizability testing, and permeability ablation experiments, thereby validating the proposed strategy's effectiveness. Compared to the baseline, our strategy increases throughput by 10% and decreases energy use by 9%. Specifically, increasing the penetration rate of CAVs significantly enhances traffic throughput, though it also increases energy consumption.