# Zero-shot Deep Reinforcement Learning Driving Policy Transfer for Autonomous Vehicles based on Robust Control

arXiv: https://arxiv.org/abs/1812.03216 · HTML: https://mlsanity.com/paper/1812.03216/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 Dec 2018.

## Built on by (most cited first, 7 of 7)

- [Mobility VLA: Multimodal Instruction Navigation with Long-Context VLMs and Topological Graphs](https://mlsanity.com/paper/2407.07775/) (2024)
- [Hierarchical Program-Triggered Reinforcement Learning Agents For Automated Driving](https://mlsanity.com/paper/2103.13861/) (2021)
- [SURE-Val: Safe Urban Relevance Extension and Validation](https://mlsanity.com/paper/2308.02266/) (2023)
- [Weakly Supervised Reinforcement Learning for Autonomous Highway Driving via Virtual Safety Cages](https://mlsanity.com/paper/2103.09726/) (2021)
- [History Encoding Representation Design for Human Intention Inference](https://mlsanity.com/paper/2106.02222/) (2021)
- [Error-Aware Policy Learning: Zero-Shot Generalization in Partially Observable Dynamic Environments](https://mlsanity.com/paper/2103.07732/) (2021)
- [Conservative Estimation of Perception Relevance of Dynamic Objects for Safe Trajectories in Automotive Scenarios](https://mlsanity.com/paper/2307.10873/) (2023)

## Abstract (arXiv)

Although deep reinforcement learning (deep RL) methods have lots of strengths that are favorable if applied to autonomous driving, real deep RL applications in autonomous driving have been slowed down by the modeling gap between the source (training) domain and the target (deployment) domain. Unlike current policy transfer approaches, which generally limit to the usage of uninterpretable neural network representations as the transferred features, we propose to transfer concrete kinematic quantities in autonomous driving. The proposed robust-control-based (RC) generic transfer architecture, which we call RL-RC, incorporates a transferable hierarchical RL trajectory planner and a robust tracking controller based on disturbance observer (DOB). The deep RL policies trained with known nominal dynamics model are transfered directly to the target domain, DOB-based robust tracking control is applied to tackle the modeling gap including the vehicle dynamics errors and the external disturbances such as side forces. We provide simulations validating the capability of the proposed method to achieve zero-shot transfer across multiple driving scenarios such as lane keeping, lane changing and obstacle avoidance.