# Reinforcement Learning

> 强化学习技能集合 - 策略梯度、值函数、模型-based、多智能体、超参数调优、ROS2集成

- Skill: `miuav/reinforcement-learning` (Agent Skill)
- Install (CLI): `npx skillmds@latest add miuav/reinforcement-learning`
- Raw SKILL.md: https://api.skillmd.com/api/skills/miuav/reinforcement-learning/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: miuav (https://skillmd.com/u/miuav)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/miuav/reinforcement-learning

---


# 强化学习技能

> 机器人强化学习完整技能集 - 从基础算法到ROS2部署

---

## 子技能

| 技能 | 描述 |
|---|---|
| rl-policy-gradient | 策略梯度方法 PPO、A3C、SAC |
| rl-value-based | DQN、Double DQN、PER 等值函数方法 |
| rl-model-based | World Models、MPC 等模型-based RL |
| rl-multi-agent | 多智能体强化学习协同 |
| rl-hyperparameter-tuning | 超参数自动调优 |
| rl-ros2-integration | ROS2 环境集成与部署 |