---
title: "CORAL：基于LiDAR的城市目标导向驾驶课程优化奖励自适应方法"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2608.14332v1"
permalink: "https://feed.astralor.com/items/itm_948a17a3d0fa0554af32ad11"
published: "2026-08-14T14:22:05.000Z"
score: 72
channel: "tech"
tags: ["CORAL","LiDAR","PPO","奖励自适应","自动驾驶","课程学习"]
---

# CORAL：基于LiDAR的城市目标导向驾驶课程优化奖励自适应方法

## AI 点评

> 这篇工作的巧妙之处在于把“学什么”和“怎么奖”同步调度，而不是单纯堆网络结构。对做自动驾驶决策或机器人导航的团队，值得借鉴其课程设计和奖励权重变化的思路，尤其是它用很低维的状态证明了泛化性。

## 速览

CORAL提出课程优化奖励自适应，用于LiDAR目标导向城市驾驶。它并行推进五阶段课程与阶段感知奖励，前者延长路线并收紧约束，后者将奖励权重从任务进度转向路线跟随、安全、平顺与规则。策略为多流Actor-Critic，在CARLA用PPO训练，仅用99维状态。相比两个PPO基线，最长路线全约束下CORAL二十次评估全部到达目标，基线仅5%和10%；消融显示缺任一机制性能下降，禁用则成功率降至55%。单城训练后零样本迁移至七个未见城镇，成功率68-98%，横向偏差低于0.35m。

## 关键点

- 提出CORAL，将五阶段课程与阶段感知奖励联合调度，解决固定奖励下长时程驾驶行为学习顺序问题。
- 仅用99维状态（极坐标LiDAR直方图+车辆遥测+路线几何+交通规则指示），无需点云编码器或BEV。
- 在最长路线、全约束条件下达到100%成功率，两个PPO基线仅5%和10%；消融证明两种机制结合的必需性。
- 单城训练后可零样本迁移至七个未见城镇，成功率68-98%，平均横向偏差低于0.35米。

## 全文

Reinforcement learning is promising for autonomous urban driving, but long-horizon goal-directed navigation asks a policy to acquire several competing behaviors at once--reaching a distant goal, tracking a route, avoiding obstacles, obeying signals--and a fixed objective gives no order in which to learn them. This paper presents CORAL, which advances two schedules together: a five-stage curriculum that progressively lengthens routes and tightens behavioral constraints, and a stage-aware reward whose component weights shift emphasis from mission progress toward route following, safety, smoothness, and rule compliance as the task hardens. The policy is a multi-stream actor-critic network trained with Proximal Policy Optimization (PPO) in CARLA on a compact 99-dimensional state pairing a polar LiDAR histogram with vehicle telemetry, ego-frame route geometry, and traffic-rule indicators--no point-cloud encoder, no bird's-eye-view rasterization. Against two PPO baselines under an identical protocol, CORAL reaches the goal in all twenty evaluation episodes on the longest routes under the full set of behavioral constraints, where the baselines reach 5% and 10%; a factorial ablation shows that neither schedule alone matches their combination: removing either lowers both success and route completion, and disabling both drops success to 55%. Trained in one town, the policy transfers zero-shot to seven unseen towns, succeeding in 68-98% of episodes on routes of the same 100-150 m length, with mean lateral deviation below 0.35 m.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_948a17a3d0fa0554af32ad11