Skip to content

AmireuxOuO/Constrained_Kinodynamic_Trajectory_Optimization

Repository files navigation

Constrained Kinodynamic Trajectory Optimization

Optimal control problems (OCP) are discretized and formulated as nonlinear programming (NLP) problems.

Dependencies

To get started, ensure the following dependencies are already installed:

  • CasADi: Symbolic automatic differentiation and optimization.
  • Pinocchio3: Library for both numerical and symbolic rigid body kinematics and dynamics.
  • CuRobo: Collision geometry generation.

Demo

Algorithms

  • Direct discretization of optimal control problems using Gauss-Legendre collocation method.
  • Constrained dynamics on configuration manifolds using local parametrization on targent charts.

The repository includes demo implementations for the following robotic systems:

  • Simple unconstrained systems (with collision avoidance) -> UR demo
  • Systems with holonomic constraints -> five-bar parallel robot demo

TODO

  • RRT path planning for generating initial guesses in NLP with collision avoidance.
  • General collision geometries, i.e., cuboids, cylinders, meshes, etc.
  • Systems with non-holonomic constriants.

Reference

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published