arXiv subjects

Shiyue Zhao

Publications and source records attributed to Shiyue Zhao.

7 recordsLinked to original sources

arXiv2026

Hamilton-Jacobi Reachability-Based Safe Reinforcement Learning for Emergency Collision Avoidance

Emergency collision avoidance under extreme driving conditions demands safety-critical control that accounts for both obstacle proximity and vehicle dynamic stability over a future time horizon, yet existing methods often rely on instantaneous or local safety evaluations. This paper proposes a safe reinforcement learning framework guided by a Hamilton-Jacobi (HJ) reachability based motion safety set that provides forward-looking safety supervision for constrained policy optimization. Specifically, a unified signed safety function is formulated by combining geometric collision margins and chassis stability limits, and is then extended through reachability analysis into a finite-horizon motion safety set that characterizes whether safety can be maintained under future vehicle state evolution. To enable practical computation, the motion safety set is approximated from offline extreme driving data, mitigating the computational burden of grid-based HJ solvers. The learned motion safety set is then embedded as a continuous safety cost into a constrained Markov decision process, and a PID-Lagrangian policy optimization scheme is employed to adaptively regulate the Lagrange multiplier for safety constraint enforcement. Simulation and real-vehicle experiments on low-adhesion obstacle-avoidance scenarios demonstrate that the proposed method achieves higher goal-reaching rates, produces smoother avoidance maneuvers, and maintains larger unified safety margins than baseline methods.

eess.SY↗

arXiv2025

RiskNet: Interaction-Aware Risk Forecasting for Autonomous Driving in Long-Tail Scenarios

Ensuring the safety of autonomous vehicles (AVs) in long-tail scenarios remains a critical challenge, particularly under high uncertainty and complex multi-agent interactions. To address this, we propose RiskNet, an interaction-aware risk forecasting framework, which integrates deterministic risk modeling with probabilistic behavior prediction for comprehensive risk assessment. At its core, RiskNet employs a field-theoretic model that captures interactions among ego vehicle, surrounding agents, and infrastructure via interaction fields and force. This model supports multidimensional risk evaluation across diverse scenarios (highways, intersections, and roundabouts), and shows robustness under high-risk and long-tail settings. To capture the behavioral uncertainty, we incorporate a graph neural network (GNN)-based trajectory prediction module, which learns multi-modal future motion distributions. Coupled with the deterministic risk field, it enables dynamic, probabilistic risk inference across time, enabling proactive safety assessment under uncertainty. Evaluations on the highD, inD, and rounD datasets, spanning lane changes, turns, and complex merges, demonstrate that our method significantly outperforms traditional approaches (e.g., TTC, THW, RSS, NC Field) in terms of accuracy, responsiveness, and directional sensitivity, while maintaining strong generalization across scenarios. This framework supports real-time, scenario-adaptive risk forecasting and demonstrates strong generalization across uncertain driving environments. It offers a unified foundation for safety-critical decision-making in long-tail scenarios.

cs.RO↗

arXiv2025

SACA: A Scenario-Aware Collision Avoidance Framework for Autonomous Vehicles Integrating LLMs-Driven Reasoning

Reliable collision avoidance under extreme situations remains a critical challenge for autonomous vehicles. While large language models (LLMs) offer promising reasoning capabilities, their application in safety-critical evasive maneuvers is limited by latency and robustness issues. Even so, LLMs stand out for their ability to weigh emotional, legal, and ethical factors, enabling socially responsible and context-aware collision avoidance. This paper proposes a scenario-aware collision avoidance (SACA) framework for extreme situations by integrating predictive scenario evaluation, data-driven reasoning, and scenario-preview-based deployment to improve collision avoidance decision-making. SACA consists of three key components. First, a predictive scenario analysis module utilizes obstacle reachability analysis and motion intention prediction to construct a comprehensive situational prompt. Second, an online reasoning module refines decision-making by leveraging prior collision avoidance knowledge and fine-tuning with scenario data. Third, an offline evaluation module assesses performance and stores scenarios in a memory bank. Additionally, A precomputed policy method improves deployability by previewing scenarios and retrieving or reasoning policies based on similarity and confidence levels. Real-vehicle tests show that, compared with baseline methods, SACA effectively reduces collision losses in extreme high-risk scenarios and lowers false triggering under complex conditions. Project page: https://sean-shiyuez.github.io/SACA/.

cs.RO↗

arXiv2025

Safety-Critical Human-Machine Shared Driving for Vehicle Collision Avoidance based on Hamilton-Jacobi reachability

Road safety continues to be a pressing global issue, with vehicle collisions imposing significant human, societal, and economic burdens. Human-machine shared collision avoidance in critical collision scenarios aims to aid drivers' accident avoidance through intervening only when necessary. Existing methods count on replanning collision-free trajectories and imposing human-machine tracking, which usually interrupts the driver's intent and increases the risk of conflict. This paper introduces a Reachability-Aware Reinforcement Learning (RL) framework for shared control, guided by Hamilton-Jacobi (HJ) reachability analysis. Machine intervention is activated only when the vehicle approaches the Collision Avoidance Reachable Set (CARS), which represents states where collision is unavoidable. First, we precompute the reachability distributions and the CARS by solving the Bellman equation using offline data. To reduce human-machine conflicts, we develop a driver model for sudden obstacles and propose an authority allocation strategy considering key collision avoidance features. Finally, we train a RL agent to reduce human-machine conflicts while enforcing the hard constraint of avoiding entry into the CARS. The proposed method was tested on a real vehicle platform. Results show that the controller intervenes effectively near CARS to prevent collisions while maintaining improved original driving task performance. Robustness analysis further supports its flexibility across different driver attributes.

cs.RO↗

arXiv2024

SafeDrive: Knowledge- and Data-Driven Risk-Sensitive Decision-Making for Autonomous Vehicles with Large Language Models

Recent advancements in autonomous vehicles (AVs) use Large Language Models (LLMs) to perform well in normal driving scenarios. However, ensuring safety in dynamic, high-risk environments and managing safety-critical long-tail events remain significant challenges. To address these issues, we propose SafeDrive, a knowledge- and data-driven risk-sensitive decision-making framework to enhance AV safety and adaptability. The proposed framework introduces a modular system comprising: (1) a Risk Module for quantifying multi-factor coupled risks involving driver, vehicle, and road interactions; (2) a Memory Module for storing and retrieving typical scenarios to improve adaptability; (3) a LLM-powered Reasoning Module for context-aware safety decision-making; and (4) a Reflection Module for refining decisions through iterative learning. By integrating knowledge-driven insights with adaptive learning mechanisms, the framework ensures robust decision-making under uncertain conditions. Extensive evaluations on real-world traffic datasets, including highways (HighD), intersections (InD), and roundabouts (RounD), validate the framework's ability to enhance decision-making safety (achieving a 100% safety rate), replicate human-like driving behaviors (with decision alignment exceeding 85%), and adapt effectively to unpredictable scenarios. SafeDrive establishes a novel paradigm for integrating knowledge- and data-driven methods, highlighting significant potential to improve safety and adaptability of autonomous driving in high-risk traffic scenarios. Project Page: https://mezzi33.github.io/SafeDrive/

cs.AI↗

arXiv2024

High-Speed Cornering Control and Real-Vehicle Deployment for Autonomous Electric Vehicles

Executing drift maneuvers during high-speed cornering presents significant challenges for autonomous vehicles, yet offers the potential to minimize turning time and enhance driving dynamics. While reinforcement learning (RL) has shown promising results in simulated environments, discrepancies between simulations and real-world conditions have limited its practical deployment. This study introduces an innovative control framework that integrates trajectory optimization with drift maneuvers, aiming to improve the algorithm's adaptability for real-vehicle implementation. We leveraged Bezier-based pre-trajectory optimization to enhance rewards and optimize the controller through Twin Delayed Deep Deterministic Policy Gradient (TD3) in a simulated environment. For real-world deployment, we implement a hybrid RL-MPC fusion mechanism, , where TD3-derived maneuvers serve as primary inputs for a Model Predictive Controller (MPC). This integration enables precise real-time tracking of the optimal trajectory, with MPC providing corrective inputs to bridge the gap between simulation and reality. The efficacy of this method is validated through real-vehicle tests on consumer-grade electric vehicles, focusing on drift U-turns and drift right-angle turns. The control outcomes of these real-vehicle tests are thoroughly documented in the paper, supported by supplementary video evidence (https://youtu.be/5wp67FcpfL8). Notably, this study is the first to deploy and apply an RL-based transient drift cornering algorithm on consumer-grade electric vehicles.

cs.RO↗

arXiv2024

Adaptive Decision-Making for Autonomous Vehicles: A Learning-Enhanced Game-Theoretic Approach in Interactive Environments

This paper proposes an adaptive behavioral decision-making method for autonomous vehicles (AVs) focusing on complex merging scenarios. Leveraging principles from non-cooperative game theory, we develop a vehicle interaction behavior model that defines key traffic elements and integrates a multifactorial reward function. Maximum entropy inverse reinforcement learning (IRL) is employed for behavior model parameter optimization. Optimal matching parameters can be obtained using the interaction behavior feature vector and the behavior probabilities output by the vehicle interaction model. Further, a behavioral decision-making method adapted to dynamic environments is proposed. By establishing a mapping model between multiple environmental variables and model parameters, it enables parameters online learning and recognition, and achieves to output interactive behavior probabilities of AVs. Quantitative analysis employing naturalistic driving datasets (highD and exiD) and real-vehicle test data validates the model's high consistency with human decision-making. In 188 tested interaction scenarios, the average human-like similarity rate is 81.73%, with a notable 83.12% in the highD dataset. Furthermore, in 145 dynamic interactions, the method matches human decisions at 77.12%, with 6913 consistence instances. Moreover, in real-vehicle tests, a 72.73% similarity with 0% safety violations are obtained. Results demonstrate the effectiveness of our proposed method in enabling AVs to make informed adaptive behavior decisions in interactive environments.

cs.MA↗