浙大高飞课题组
微分平坦
微分平坦的思想是:一个全维度的状态空间可以被一组低维的精心挑选的输出平坦空间(flat-output space)的变量及其导数的代数组合的方式所表示。由此,轨迹规划就可以在这组精心挑选的变量的空间所进行。
论文:An Efficient Spatial-Temporal Trajectory Planner for Autonomous Vehicles
开源项目:
ZJU-FAST-Lab/Dftpav:基于微分平整度的轻量级类车机器人轨迹规划仪 (github.com)
知乎论文讲解:
An Efficient Spatial-Temporal Trajectory Planner(基于微分平坦轨迹规划) - 知乎 (zhihu.com)
港科大沈劭劼课题组
MPDM
论文:Multipolicy Decision-Making in Dynamic, Uncertain Environments for Autonomous Driving
论文:Multipolicy Decision-Making for Autonomous Driving via Changepoint-based Behavior Prediction
论文:Multipolicy decision-making for autonomous driving via changepoint-based behavior prediction: Theory and experiment
Revisit Multi-policy Decision-making:MPDM, EUDM, EPSILON(Part 1) - 知乎 (zhihu.com)
EUDM
论文:Efficient Uncertainty-aware Decision-making for Automated Driving Using Guided Branching
EPSILON
论文:EPSILON: An Efficient Planning System for Automated Vehicles in Highly Interactive Environments
开源项目:HKUST-Aerial-Robotics/EPSILON (github.com)
Revisit Multi-policy Decision-making:MPDM, EUDM, EPSILON(Part 1) - 知乎 (zhihu.com)
MARC
论文:MARC: Multipolicy and Risk-aware Contingency Planning for Autonomous Driving
未开源
论文讲解:MARC: Multipolicy and Risk-aware Contingency Planning for Autonomous Driving - 知乎 (zhihu.com)
湖大李柏课题组
笛卡尔坐标下规划
论文:Autonomous Driving on Curvy Roads Without Reliance on Frenet Frame: A Cartesian-Based
Trajectory Planning Method
论文讲解:Autonomous Driving on Curvy Roads without frenet - 知乎 (zhihu.com)
李柏老师关于泊车的也有很多开源的项目,这里就不再列举了,可以去github看看。
其他课题组
使用蒙特卡洛树搜索 (MCTS) 的自动驾驶决策
论文:Monte-Carlo Tree Search for Behavior Planning in Autonomous Driving
开源项目:中顺/MCTS_for_Behavior_Planning (github.com)
论文讲解:Monte-Carlo Tree Search for Behavior Planning in Autonomous Driving - 知乎 (zhihu.com)
处理Y字型合流场景下的目标车合作可能性估计问题
论文:Cooperation-Aware Reinforcement Learning for Merging in Dense Traffic
开源项目:sisl/AutonomousMerging.jl: Implementation of a highway merging scenario (github.com)
论文讲解:Cooperation-Aware Reinforcement Learning for Merging in Dense Traffic - 知乎 (zhihu.com)
TODO 持续更新