CV
Curriculum Vitae of Zhihao Lin — Ph.D. in Autonomous Systems & Connectivity, University of Glasgow.
Contact Information
| Name | Zhihao Lin |
| Professional Title | Ph.D. in Autonomous Systems & Connectivity |
| kk43327897@gmail.com |
Professional Summary
Ph.D. researcher in Autonomous Systems & Connectivity at the University of Glasgow. Research on structure-aware reinforcement learning, multi-agent decision making, and perception-integrated autonomy. On Google Scholar: h-index of 11 and over 500 citations (as of May 2026).
Experience
-
2025 - Glasgow, UK
Doctoral Researcher, University of Glasgow
Structure-Aware Learning for Policy, Reasoning, and Decision
- Developed advantage-guided and preference-based methods for adaptive exploration in RL.
- Designed stability-aware objectives and proxy supervision signals to improve structural generalization.
- Formulated geometry-aligned policy learning principles beyond Gaussian action distributions.
- Built unified perspectives linking action manifolds, Bellman stability, and structural RL supervision.
-
2024 - 2025 Glasgow, UK & France
Doctoral Researcher, UofG & LAMIH
Multi-Agent Safety Planning in Dynamic Scenarios
Supervisors: Dr. J. Lan, Prof. D. Flynn, Dr. Christos Anagnostopoulos, Dr. A. Nguyen
- Developed MCTS-based coordination algorithms for interactive driving at unsignalized intersections.
- Proposed a scalable framework for mixed traffic with centralized and decentralized policies.
- Developed safety-critical pruning and interaction-graph filtering to reduce planning complexity.
- Designed contingency-aware trajectory optimization for safe planning under uncertainty.
-
2023 - 2024 Glasgow, UK
Doctoral Researcher, University of Glasgow
Interactive Decision Making for Autonomy
Supervisors: Dr. Jianglin Lan and Prof. Dezong Zhao
- Developed KAN-based RL systems for driving tasks like roundabouts and ramp merging.
- Proposed multi-agent actor-critic models with LSTM intention modeling for road interactions.
- Designed reward shaping and exploration for efficient policy learning in racing scenarios.
- Integrated uncertainty-aware game-theoretic planning modules for safe interactive autonomy.
-
2023 - 2023 Glasgow, UK
Doctoral Researcher, UofG & SJTU
Robust and Dynamic-Aware SLAM for Autonomy
Supervisors: Dr. Jianglin Lan and Dr. Hanyang Zhuang
- Developed visual SLAM for dynamic environments using point-line features and semantic cues.
- Built a multi-modal mapping pipeline that filters dynamic objects via detection and epipolar constraints.
- Integrated SLAM with CLF-CBF and TEB planning for safe navigation in unstructured scenes.
- Optimized trajectories with stability and dynamic-awareness for long-term autonomous deployment.
Education
-
2023 - 2026 Glasgow, UK
Ph.D.
University of Glasgow
Autonomous Systems & Connectivity
- Supervisor: Dr. Jianglin Lan
- Research interests: Structure-Aware RL, Multi-Agent Decision Making, and Perception-Integrated Autonomy
-
2025 - 2025 Valenciennes, France
Visiting Research
Université Polytechnique Hauts-de-France (LAMIH UMR CNRS 8201)
Control Department — Visiting Research Student
- Host supervisor: Dr. Anh-Tu Nguyen
-
2019 - 2022 Changchun, China
M.S.
Jilin University
Circuits and Systems
-
2014 - 2018 Liaocheng, China
B.S.
Liaocheng University
Electronic Information Engineering
Honors and Awards
-
2023 Doctoral Scholarship
China Scholarship Council (CSC)
2023–2026
-
2025 Research Mobility Fund
University of Glasgow
£1,800 for international research collaboration.
-
2025 Editor's Choice Award
Sensors Journal
For “Enhanced Visual SLAM for Collision-Free Driving with Lightweight Autonomous Cars”.
-
2021 Academic Scholarship
Jilin University
2020–2021
Open Source Projects
-
GAC — Geometric Action Control
Official implementation of Beyond Distributions: Geometric Action Control for Continuous Reinforcement Learning (ICLR 2026).
-
AMS — Action Manifold Smoothing
Official implementation of Action Manifold Smoothing: A Lipschitz Pathway Perspective on High-Dimensional Reinforcement Learning (ICML 2026).
Invited Talks and Presentations
- Conference Presentation — “Action Manifold Smoothing,” ICML 2026, Seoul, South Korea (July 2026)
- Conference Presentation — “Reconstructed Level-k MCTS,” AAMAS 2026 (May 2026)
- Conference Presentation — “Geometric Action Control,” ICLR 2026 (Apr 2026)
- Invited Talk — “Geometric Action Control,” KTH Royal Institute of Technology, Stockholm (Jan 2026)
- Invited Research Visit & Presentation — Université Polytechnique Hauts-de-France (LAMIH UMR CNRS 8201), France (Nov 2025)
- Conference Presentation — “Distributed MCTS for Safe Coordination,” IEEE ICDCS 2025 (Jul 2025)
Skills
Programming Languages: Python, C/C++, MATLAB, R, LaTeX
Machine Learning Tools: PyTorch, Scikit-learn, Pandas, NumPy
Academic Service
- Reviewer, IEEE Transactions — T-RO, T-ITS, T-IV, T-ASE, T-TE, T-CSVT, T-IM
- Reviewer, Journals — Transp. Res. Part C (TR-C), Journal of Field Robotics (JFR), Pattern Recognition, Neural Networks, Neurocomputing, ISA Transactions
Languages
Chinese : Native speaker
English : Fluent