无人驾驶路径规划笔记
无人驾驶路径规划笔记
企业JD要求：
路径规划算法：A*, RRT, NMPC, MPC, lattice
凸优化： QP, SQP
深度／强化学习：Paddle, Tensorflow2
马尔科夫决策过程
马尔科夫特性(Markov Property) 当且仅当满足下式，状态 S t S_t St具有马尔科夫特性
P [ S t + 1 ∣ S t ] = P [ S t + 1 ∣ S 1 , . . . , S t ] P[S_{t+1}|S_t] = P[S_{t+1}| S_1,...,S_t] P[St+1∣St]=P[St+1∣S1,...,St]状态转移概率(State Transition Probability)对于马尔科夫状态s和下一状态s’状态转移概率可定义为： P s s ′ = P [ S t + 1 = s ′ ∣ S t = s ] P_{ss'}=P[S_{t+1}=s'|S_{t}=s] Pss′=P[St+1=s′∣St=s]马尔科夫过程(Markov Process)
是一个无记忆的随机过程
马尔科夫决策过程(Markov Decision Process, MDP)是增加了动作选择的马尔科夫过程
贝尔曼最优公式(Bellman Optimality Equation):…
Constraints 3 类
Local constraints: e.g. avoid collision with obstacles
Differential constrains: e.g. bounded curvature
Global constraints
Path-Planning Approaches
- Roadmap
- Cell decomposition
- Potential field
Lattice Planning
决策规划分类：
Rules based -> Optimization -> Data based
优化三个要素：1. Objective function 2.Constraint 3…
路径追踪
1, Pure pursuit