Apollo 自动驾驶工程师技能图谱

优质

小牛编辑

137浏览

2023-12-01

本图谱核心贡献者：Apollo 技术布道师团队

基于 Apollo 技术框架的自动驾驶工程师技能图谱：跟公布的技术框架结构保持一致，强调了 Apollo 软件之外的硬件能力。

Toolbox（Basic）

Localization
Perception
- Deep Learning:CNN
- Low-level Vision
- 图形学与视觉几何
- 状态估计与滤波
- 最优化问题求解
Planning
- Smoothing Curves
- Vehicle Kinematics
- Station-time 2D Model
- Optimization Algorithms:Dynamic Programming
- Quadratic Programming
- Smoothing Spline
- A*
- Probability Math Models:Bayesian
- Markov Process
- MDP
- Machine Learning:
  - CNN
  - RNN
  - Reinforcement Learning
- Motion Planning Background:Sampling Based Motion Planning
- Optimization Based Motion Planning
- ComputerScience:Parallel Programming
- For Planning Focused:Basic Motion Planning Technique
- Dynamic Programming
- Computation Geometry
- 深度学习算法、BP 算法
Control
- Basic Control Theory
- Linear-Quadratic Regulator (LQR)
- Model Predictive Control(MPC)
- PID Basic Control Theory
End-to-End
- Deep Learning
- CV
- 逆向汽车动力学模型
- 端到端驾驶模型：
  - 横向模型：CNN
  - 纵向模型：Convolutional-LSTM
  - 横 + 纵向模型：LRCN