Curriculum-Driven Deep Reinforcement Learning for Autonomous Far-Range Rocket Landing
Paper · https://arxiv.org/abs/2610.35954
arXiv:2610.35954v1 Announce Type: new Abstract: Autonomous vertical landing of reusable launch vehicles is a critical yet formidable challenge in aerospace engineering, primarily due to highly nonlinear dynamics, stringent precision require- ments, and inherent environmental uncertainties. While reinforcement learning (RL) has emerged as a promising alternative to traditional model-based control, existing approaches frequently rely on flat training paradigms and highly simplified initial conditions, drastically limiting their real-world applicability. This paper proposes a comprehensive curriculum-driven deep reinforcement learning framework designed to master both far-range navigation and precision landing. By employing a three-stage curriculum augmented with a critic warm-up procedure and a rehearsal mechanism to prevent catastrophic forgetting, alongside a high-dimensional observation vector utilizing a full 3 X 3 rotation-matrix, the agent is trained to navigate from an initial horizontal distance of 1,000 to 1,500 units vastly farther than prior benchmarks before executing a soft touchdown. Crucially, we intro- duce a redesigned reward function that reduces unnecessary computational overhead and addresses limitations of earlier formulations, including poorly calibrated reward components and potentially conflicting optimization objectives. The redesigned function bounds the per-step reward magnitude by (O(1)), while its progress term accumulates the total distance closed over the trajectory. The available 100,000-episode results report a 99.98% success rate for the proposed policy and 68.31% for the baseline under the evaluation conditions. In the evaluation, the proposed policy retained 69.1% of its initial fuel capacity and achieved a mean final distance from the pad centre of 7.07 units, compared with 13.10 units for the baseline. Landing visualizations are available in Video 1 and Video 2.
First seen by Oriel on .
AI summary
This paper presents a three-stage deep reinforcement learning curriculum for autonomous far-range rocket navigation and precision landing, reporting substantially better success, fuel retention, and touchdown accuracy than a baseline.
- The curriculum includes critic warm-up and rehearsal to prevent catastrophic forgetting, with observations represented using a full 3 X 3 rotation-matrix.
- The agent learns from initial horizontal distances of 1,000 to 1,500 units before executing a soft touchdown.
- The redesigned reward function bounds per-step reward magnitude by O(1), removes conflicting or poorly calibrated components, and accumulates total distance closed.
- Across 100,000 episodes, the proposed policy achieved 99.98% success under evaluation conditions, compared with 68.31% for the baseline.
- The policy retained 69.1% of initial fuel and averaged 7.07 units from the pad centre, versus 13.10 units for the baseline.
#deep reinforcement learning #curriculum learning #autonomous landing #aerospace
Generated by gpt-5.6-sol. Check the original for details.
Aliases
| Source | External ID | URL |
|---|---|---|
| arXiv AI Papers | oai:arXiv.org:2610.35954v1 | https://arxiv.org/abs/2610.35954 Read discussion |