A tree-based randomized planner that extends the roadmap by sampling the neighborhood of existing samples. More...

Inheritance diagram for PerturbationTreePlanner:

Public Member Functions
	PerturbationTreePlanner (CSpace *s)

virtual void	GenerateConfig (Config &x)

virtual Node *	AddMilestone (const Config &x)

virtual void	Cleanup ()

virtual Node *	SelectMilestone (const std::vector< Node * > &milestones)

Public Member Functions inherited from TreeRoadmapPlanner
	TreeRoadmapPlanner (CSpace *)

virtual Node *	TestAndAddMilestone (const Config &x)

virtual Node *	AddInfeasibleMilestone (const Config &x)

virtual Node *	Extend ()

virtual void	ConnectToNeighbors (Node *)

virtual EdgePlannerPtr	TryConnect (Node , Node )

virtual void	DeleteSubtree (Node *n)

virtual Node *	ClosestMilestone (const Config &x)

virtual int	ClosestMilestoneIndex (const Config &x)

virtual Node *	ClosestMilestoneInComponent (int component, const Config &x)

virtual Node *	ClosestMilestoneInSubtree (Node *node, const Config &x)

Node *	Extend (Node *n, const Config &x)

Node *	TryExtend (Node *n, const Config &x)

void	AttachChild (Node p, Node c, const EdgePlannerPtr &e)

Node *	SplitEdge (Node p, Node n, Real u)

void	CreatePath (Node a, Node b, MilestonePath &path)
	Creates the unique path from a to b. These MUST be in the same connected component.

virtual Real	OptimizePath (Node a, const std::vector< Node > &goals, ObjectiveFunctionalBase *cost, MilestonePath &path)

Public Attributes
Real	delta
	Neighborhood distance.

std::vector< Real >	weights
	Node selection weights.

Public Attributes inherited from TreeRoadmapPlanner
CSpace *	space

std::vector< Node * >	connectedComponents

Real	connectionThreshold

std::vector< Vector >	milestoneConfigs

std::vector< Node * >	milestones

std::shared_ptr< PointLocationBase >	pointLocator

Config	x

Additional Inherited Members
Public Types inherited from TreeRoadmapPlanner
typedef Graph::TreeNode< Milestone, EdgePlannerPtr >	Node

Detailed Description

A tree-based randomized planner that extends the roadmap by sampling the neighborhood of existing samples.

The existing sample picked is selected with probability proportion to its value in the weight vector. By default, a new node is given weight 1.

The documentation for this class was generated from the following files:

Public Member Functions