用可微仿真提升自动驾驶路径规划精度
Autonomous Vehicle Path Planning by Searching With Differentiable Simulation
- 基于可微仿真器Waymax进行动作序列搜索
- 相比传统方法，路径跟踪与规划准确率显著提升
- 适合需要高安全性的自动驾驶场景
规划使智能体能在真实世界执行前安全地优化动作。在自动驾驶中，这至关重要，可避免碰撞并在复杂密集的交通环境中导航。一种规划方式是搜索最佳动作序列，但当策略、状态预测器和评判器均需学习时，这一过程极具挑战。本文提出可微仿真搜索（DSS）框架，利用可微仿真器Waymax作为状态预测器和评判器。该框架依赖仿真器的硬编码动力学，实现高精度状态预测，并借助其可微性对动作序列进行高效搜索。DSS智能体通过梯度下降优化想象中的未来轨迹。实验表明，结合规划梯度与随机搜索的DSS显著优于序列预测、模仿学习、无模型强化学习及其他规划方法，在路径跟踪与规划准确性上均有明显提升。
原文摘要 · Abstract (English)
Planning allows an agent to safely refine its actions before executing them in the real world. In autonomous driving, this is crucial to avoid collisions and navigate in complex, dense traffic scenarios. One way to plan is to search for the best action sequence. However, this is challenging when all necessary components - policy, next-state predictor, and critic - have to be learned. Here we propose Differentiable Simulation for Search (DSS), a framework that leverages the differentiable simulator Waymax as both a next state predictor and a critic. It relies on the simulator's hardcoded dynamics, making state predictions highly accurate, while utilizing the simulator's differentiability to effectively search across action sequences. Our DSS agent optimizes its actions using gradient descent over imagined future trajectories. We show experimentally that DSS - the combination of planning gradients and stochastic search - significantly improves tracking and path planning accuracy compared to sequence prediction, imitation learning, model-free RL, and other planning methods.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。