Detailed Description

Options for finding polytope with given Lagrangians.

#include <drake/geometry/optimization/cspace_free_polytope.h>

Public Attributes
std::optional< double >	backoff_scale {std::nullopt}

double	ellipsoid_margin_epsilon {1E-5}
	We will maximize the cost ∏ᵢ (δᵢ + ε) where δᵢ is the margin from each face of the polytope {s \| Cs<=d} to the inscribed ellipsoid, ε is ellipsoid_margin_epsilon, a small positive constant to make sure δᵢ + ε being strictly positive. More...

solvers::SolverId	solver_id {solvers::MosekSolver::id()}
	ID for the solver. More...

std::optional< solvers::SolverOptions >	solver_options {std::nullopt}
	options for solving the MathematicalProgram More...

std::optional< Eigen::MatrixXd >	s_inner_pts
	We can constrain the C-space polytope {s \| C*s<=d, s_lower<=s<=s_upper} to contain some sampled s. More...

bool	search_s_bounds_lagrangians {true}
	If set to true, then we will also search for the Lagrangian multipliers for the constraint s_lower <= s <= s_upper; otherwise we fix the Lagrangian multiplier to the solution found when we fix the C-space polytope {s \| C*s<=d, s_lower<=s<=s_upper}. More...

EllipsoidMarginCost	ellipsoid_margin_cost
	Type of cost on the ellipsoid margin. More...

Member Data Documentation

◆ backoff_scale

std::optional<double> backoff_scale {std::nullopt}

◆ ellipsoid_margin_cost

EllipsoidMarginCost ellipsoid_margin_cost

Initial value:

{

EllipsoidMarginCost::kGeometricMean}

Type of cost on the ellipsoid margin.

◆ ellipsoid_margin_epsilon

double ellipsoid_margin_epsilon {1E-5}

We will maximize the cost ∏ᵢ (δᵢ + ε) where δᵢ is the margin from each face of the polytope {s | Cs<=d} to the inscribed ellipsoid, ε is ellipsoid_margin_epsilon, a small positive constant to make sure δᵢ + ε being strictly positive.

◆ s_inner_pts

std::optional<Eigen::MatrixXd> s_inner_pts

We can constrain the C-space polytope {s | C*s<=d, s_lower<=s<=s_upper} to contain some sampled s.

Each column of s_inner_pts is a sample of s.

◆ search_s_bounds_lagrangians

bool search_s_bounds_lagrangians {true}

If set to true, then we will also search for the Lagrangian multipliers for the constraint s_lower <= s <= s_upper; otherwise we fix the Lagrangian multiplier to the solution found when we fix the C-space polytope {s | C*s<=d, s_lower<=s<=s_upper}.

◆ solver_id

solvers::SolverId solver_id {solvers::MosekSolver::id()}

ID for the solver.

◆ solver_options

std::optional<solvers::SolverOptions> solver_options {std::nullopt}

options for solving the MathematicalProgram

The documentation for this struct was generated from the following file:

drake/geometry/optimization/cspace_free_polytope.h