swift_numerical

Huazhong University of Science and Technology, School of Artificial Intelligence and Automation, Control Robotics and Automation Laboratory (CoraLab) Hierarchical Planner for Agile Quadruped Navigation with Awareness of Motion Anisotropy for Numerical Calculation MATLAB code

Description

1. LazyPRM: The implementation of the LazyPRM algorithm, which mainly includes several key functions for path search as OpenList.m, CloseList.m, and Point.m.
2. LocalOBVP: LocalOBVP uses Optimal Boundary Value Problem to connect Nodes in the path search process, mainly using SampleOBVP.m.
3. map: store obstacle map.
4. PolyOpt: Nonlinear Trajectory Optimization, mainly implemented in matlatopt.m using nonlinear iterative optimization.
5. TrajGener: Piecewise curve optimization, mainly implemented in SegPolyTraj.m for polynomial and Bezier curve optimization.

Instructions

First, use LazyKinoPRM.m for initial path search. Then, use matlabopt.m for trajectory optimization to obtain a continuous and smooth spatio-temporal trajectory. Kinodynamic Trajectory Generation mapaddress: Path of the obstacle map. Point_i/Point_f: Initial point / target point represented as $[x,y,θ]$, indicating horizontal displacement and orientation angle. Nonlinear Trajectory Optimization seq_start/seq_end: Starting and ending segment indices for optimization.

Postscript

1. The MATLAB code here is the theoretical verification part of our project. This algorithm is applied in practical navigation planning.
2. If you have any questions or suggestions, please feel free to contact [email protected] or [email protected]. Thank you very much.