Zhengxian Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Zhengxian Chen has a total of 51 co-authors. Here are the top co-authors by number of shared publications:
Jun Li
5 shared publications- Chaosheng Huang
Tsinghua University
4 shared publications Boyang Xu
College of Design and Engineering, National University of Singapore, Singapore, Republic of Singapore
2 shared publicationsFojin Zhou
2 shared publications- Han Hai
School of Vehicle and Mobility, Tsinghua University, Beijing, China
2 shared publications - Huanan Wang
School of Vehicle and Mobility, Tsinghua University, Beijing, China
2 shared publications
Latest Publications
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (23) , 57887-57901
An Energy-Aware Car-Following Control Framework With Data-Driven Enhanced Deep Reinforcement Learning for Automated Vehicles
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (23) , 57887-57901
An Energy-Aware Car-Following Control Framework With Data-Driven Enhanced Deep Reinforcement Learning for Automated Vehicles
- 23 April 2026
- Published by Elsevier in International Journal of Mechanical Sciences
- Vol. 321
Spectral dynamics of corner separation controlled by blended wall contouring
- 23 April 2026
- Published by Elsevier in International Journal of Mechanical Sciences
- Vol. 321
Spectral dynamics of corner separation controlled by blended wall contouring
- 1 April 2025
- Published by AIP Publishing in Physics of Fluids
- Vol. 37 (4)
Spatio-temporal analysis of shear traction dynamics on fish body surface
Fish swimming is a prominent research focus in biomimetic hydrodynamics, and studying the surface stress dynamic of fish contributes to the shape and structural design of robotic fish as well as the energy-efficient control of schooling robotic fish. Traditional research primarily relies on conventional numerical simulations that involve mesh generation and testing. However, the deformation of a swimming fish significantly impacts local grids, and the uncertainties in mesh structure reduce simulation accuracy. To address these challenges, this study employs the open-source computational fluid dynamics solver Palabos, which is based on the lattice Boltzmann method (LBM), to investigate the stress dynamics on fish surfaces. The immersed boundary method algorithm is coupled with LBM and enhanced with a multi-direct forcing scheme to improve precision and effectively handle fluid–solid interactions. Simulations reveal the spatial variations of shear traction on the fish surface. Furthermore, dynamic mode decomposition is applied to analyze the temporal and spatial evolution of shear traction distribution. The results indicate that under steady undulation, high-speed regions and areas of high vorticity occur where the fish's head and tail make contact with the water surface. Despite changes in tail-beating frequency, the overall flow field and vortex shedding pattern remain consistent. Further stress analysis reveals that shear traction oscillates regularly, with its amplitude varying along the fish's body surface. During swimming, shear traction is concentrated at the tail tip and the fish's head edge, regions that also feature large vortices in the fluid field. Based on the shear traction distribution, it can be inferred that the primary source of propulsion originates from the fish's back. The findings provide a theoretical foundation and reference points for the design of bionic fish exoskeletons, strategies for reducing surface drag, and fluid field optimization. This study elucidates the mechanisms underlying surface shear traction distribution on fish, offering valuable insight for related research.
- 1 April 2025
- Published by AIP Publishing in Physics of Fluids
- Vol. 37 (4)
Spatio-temporal analysis of shear traction dynamics on fish body surface
Fish swimming is a prominent research focus in biomimetic hydrodynamics, and studying the surface stress dynamic of fish contributes to the shape and structural design of robotic fish as well as the energy-efficient control of schooling robotic fish. Traditional research primarily relies on conventional numerical simulations that involve mesh generation and testing. However, the deformation of a swimming fish significantly impacts local grids, and the uncertainties in mesh structure reduce simulation accuracy. To address these challenges, this study employs the open-source computational fluid dynamics solver Palabos, which is based on the lattice Boltzmann method (LBM), to investigate the stress dynamics on fish surfaces. The immersed boundary method algorithm is coupled with LBM and enhanced with a multi-direct forcing scheme to improve precision and effectively handle fluid–solid interactions. Simulations reveal the spatial variations of shear traction on the fish surface. Furthermore, dynamic mode decomposition is applied to analyze the temporal and spatial evolution of shear traction distribution. The results indicate that under steady undulation, high-speed regions and areas of high vorticity occur where the fish's head and tail make contact with the water surface. Despite changes in tail-beating frequency, the overall flow field and vortex shedding pattern remain consistent. Further stress analysis reveals that shear traction oscillates regularly, with its amplitude varying along the fish's body surface. During swimming, shear traction is concentrated at the tail tip and the fish's head edge, regions that also feature large vortices in the fluid field. Based on the shear traction distribution, it can be inferred that the primary source of propulsion originates from the fish's back. The findings provide a theoretical foundation and reference points for the design of bionic fish exoskeletons, strategies for reducing surface drag, and fluid field optimization. This study elucidates the mechanisms underlying surface shear traction distribution on fish, offering valuable insight for related research.
- Top Cited PapersOpen Access
- 19 March 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-14
Predicting driving comfort in autonomous vehicles using road information and multi-head attention models
Driving comfort is a crucial consideration in the automotive industry. In the realm of autonomous driving, comfort has always been a factor that requires continuous improvement. A common approach to improving driving comfort is through the optimization of local path planning. Nevertheless, it is imperative to recognize that macroscopic factors, including traffic flow and road conditions, wield a substantial influence on comfort. For instance, complex traffic scenarios increase the possibility of emergency braking, thereby affecting comfort. Consequently, investigating the intricate interplay between comfort and global path planning becomes essential. This paper introduces a methodology and framework for predicting driving comfort by leveraging road information. The study established a road information-driving comfort dataset and devised prediction models using multi-head attention mechanism. The ensuing discussion elucidates the practical application of the model in path planning through examples and tests. Following the path optimized by the model, the vehicles exhibited a reduction in jerk. This research predicted driving comfort based on road information and integrated it with global path planning, which holds significant implications for autonomous driving navigation systems and provides a valuable reference for related research.
- 19 March 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-14
Top Cited PapersOpen AccessPredicting driving comfort in autonomous vehicles using road information and multi-head attention models
Driving comfort is a crucial consideration in the automotive industry. In the realm of autonomous driving, comfort has always been a factor that requires continuous improvement. A common approach to improving driving comfort is through the optimization of local path planning. Nevertheless, it is imperative to recognize that macroscopic factors, including traffic flow and road conditions, wield a substantial influence on comfort. For instance, complex traffic scenarios increase the possibility of emergency braking, thereby affecting comfort. Consequently, investigating the intricate interplay between comfort and global path planning becomes essential. This paper introduces a methodology and framework for predicting driving comfort by leveraging road information. The study established a road information-driving comfort dataset and devised prediction models using multi-head attention mechanism. The ensuing discussion elucidates the practical application of the model in path planning through examples and tests. Following the path optimized by the model, the vehicles exhibited a reduction in jerk. This research predicted driving comfort based on road information and integrated it with global path planning, which holds significant implications for autonomous driving navigation systems and provides a valuable reference for related research.
- 13 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 35 (7) , 6341-6356
PDDepth: Pose Decoupled Monocular Depth Estimation for Roadside Perception System
Accurate depth information is crucial for roadside perception in Cooperative Vehicle Infrastructure Systems. Beyond existing radar and LiDAR solutions, monocular depth estimation using surveillance cameras is emerging as a superior approach due to its cost-effectiveness and dense depth output. Unlike onboard cameras, roadside cameras are relatively fixed in position. Many existing monocular depth estimation methods, which do not independently model camera pose, tend to overfit to training data and produce suboptimal results when confronted with slight variations in camera poses, which may be caused by external forces within the same camera or across different cameras. To address this issue, a pose decoupled monocular depth estimation method specifically designed for roadside perception systems is proposed. This method separates depth estimation into two components: a pose-dependent modeling portion that recovers ground depth based on the current camera pose, and a pose-agnostic portion that estimates pixel height relative to the ground plane. Additionally, a knowledge distillation framework is introduced to improve the robustness of the proposed method against variations in roadside cameras. To validate the method, we propose the first open source dataset for roadside monocular depth estimation, DAIR-MDE, and a roadside instance segmentation dataset, DAIR-Ins, both derived from the DAIR dataset. The proposed method demonstrates significant advances over the state-of-the-art methods on DAIR-MDE. The proposed dataset and source code are publicly available at https://github.com/441599828/PDDepth.
- 13 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 35 (7) , 6341-6356
PDDepth: Pose Decoupled Monocular Depth Estimation for Roadside Perception System
Accurate depth information is crucial for roadside perception in Cooperative Vehicle Infrastructure Systems. Beyond existing radar and LiDAR solutions, monocular depth estimation using surveillance cameras is emerging as a superior approach due to its cost-effectiveness and dense depth output. Unlike onboard cameras, roadside cameras are relatively fixed in position. Many existing monocular depth estimation methods, which do not independently model camera pose, tend to overfit to training data and produce suboptimal results when confronted with slight variations in camera poses, which may be caused by external forces within the same camera or across different cameras. To address this issue, a pose decoupled monocular depth estimation method specifically designed for roadside perception systems is proposed. This method separates depth estimation into two components: a pose-dependent modeling portion that recovers ground depth based on the current camera pose, and a pose-agnostic portion that estimates pixel height relative to the ground plane. Additionally, a knowledge distillation framework is introduced to improve the robustness of the proposed method against variations in roadside cameras. To validate the method, we propose the first open source dataset for roadside monocular depth estimation, DAIR-MDE, and a roadside instance segmentation dataset, DAIR-Ins, both derived from the DAIR dataset. The proposed method demonstrates significant advances over the state-of-the-art methods on DAIR-MDE. The proposed dataset and source code are publicly available at https://github.com/441599828/PDDepth.
- Open Access
- 4 January 2024
- Published by Springer Nature in Research Square
Autonomous Driving Comfort Prediction: Improve Comfort through Global Path Planning
Comfort is a pivotal consideration in the realm of autonomous driving, prompting researchers to address methods of enhancing driving comfort for autonomous driving vehicles. A central focus of current research revolves around the judicious planning of driving trajectories with a primary emphasis on comfort. Given the transient nature of comfort, prevailing path planning strategies predominantly center on local considerations. Nevertheless, it is imperative to recognize that macroscopic factors, including traffic flow and road conditions, wield a substantial influence on comfort. For instance, complex traffic scenarios elevate the likelihood of emergency braking, thereby affecting comfort. Consequently, investigating the intricate interplay between comfort and global path planning becomes essential. In the context of autonomous driving, this paper introduces a methodology and framework for predicting driving comfort by leveraging road information. The study establishes a comprehensive road information-comfort dataset and devises predictive models for driving comfort, employing a multi-head attention approach. The ensuing discussion elucidates the practical application of the model in path selection and planning through illustrative examples. Finally, real-road vehicle tests were conducted. Following the utilization of the path optimized by the model, the autonomous driving car exhibited an approximate 17.3% reduction in the mean longitudinal jerk and a 9.7% reduction in the mean lateral jerk. These results serve to validate the effectiveness and significance of the proposed model. Notably, this research marks the pioneering integration of driving comfort with global path planning, which is of immense importance for autonomous driving navigation systems, offering an insightful and valuable reference for related study.
- 4 January 2024
- Published by Springer Nature in Research Square
Open AccessAutonomous Driving Comfort Prediction: Improve Comfort through Global Path Planning
Comfort is a pivotal consideration in the realm of autonomous driving, prompting researchers to address methods of enhancing driving comfort for autonomous driving vehicles. A central focus of current research revolves around the judicious planning of driving trajectories with a primary emphasis on comfort. Given the transient nature of comfort, prevailing path planning strategies predominantly center on local considerations. Nevertheless, it is imperative to recognize that macroscopic factors, including traffic flow and road conditions, wield a substantial influence on comfort. For instance, complex traffic scenarios elevate the likelihood of emergency braking, thereby affecting comfort. Consequently, investigating the intricate interplay between comfort and global path planning becomes essential. In the context of autonomous driving, this paper introduces a methodology and framework for predicting driving comfort by leveraging road information. The study establishes a comprehensive road information-comfort dataset and devises predictive models for driving comfort, employing a multi-head attention approach. The ensuing discussion elucidates the practical application of the model in path selection and planning through illustrative examples. Finally, real-road vehicle tests were conducted. Following the utilization of the path optimized by the model, the autonomous driving car exhibited an approximate 17.3% reduction in the mean longitudinal jerk and a 9.7% reduction in the mean lateral jerk. These results serve to validate the effectiveness and significance of the proposed model. Notably, this research marks the pioneering integration of driving comfort with global path planning, which is of immense importance for autonomous driving navigation systems, offering an insightful and valuable reference for related study.