Class SPARS

Nested Relationships

Nested Types

Inheritance Relationships

Base Type

Class Documentation

class SPARS : public ompl::base::Planner

SPArse Roadmap Spanner technique.

Short description

SPARS is an algorithm which operates similarly to the Visibility-based PRM. It has several desirable properties, including asymptotic near-optimality, and a meaningful stopping criterion.

External documentation

A. Dobson, A. Krontiris, K. Bekris, Sparse Roadmap Spanners, Workshop on the Algorithmic Foundations of Robotics (WAFR) 2012. [PDF]

Public Types

enum GuardType

Enumeration which specifies the reason a guard is added to the spanner.

Values:

enumerator START
enumerator GOAL
enumerator COVERAGE
enumerator CONNECTIVITY
enumerator INTERFACE
enumerator QUALITY
using VertexIndexType = unsigned long

The type used internally for representing vertex IDs.

using InterfaceHash = std::unordered_map<VertexIndexType, std::set<VertexIndexType>>

Hash for storing interface information.

using DensePath = std::deque<base::State*>

Internal representation of a dense path.

using SpannerGraph = boost::adjacency_list<boost::vecS, boost::vecS, boost::undirectedS, boost::property<vertex_state_t, base::State*, boost::property<boost::vertex_predecessor_t, VertexIndexType, boost::property<boost::vertex_rank_t, VertexIndexType, boost::property<vertex_color_t, GuardType, boost::property<vertex_list_t, std::set<VertexIndexType>, boost::property<vertex_interface_list_t, InterfaceHash>>>>>>, boost::property<boost::edge_weight_t, base::Cost>>

The constructed roadmap spanner.

Any BGL graph representation could be used here, but the

spanner should be very sparse (m<n^2), so we use an adjacency_list.

Nodes in the spanner contain extra information needed by the

spanner technique, including nodes in the dense graph which nodes in the spanner represent.

SparseEdges should be undirected and have a weight property.

using SparseVertex = boost::graph_traits<SpannerGraph>::vertex_descriptor

A vertex in the sparse roadmap that is constructed.

using SparseEdge = boost::graph_traits<SpannerGraph>::edge_descriptor

An edge in the sparse roadmap that is constructed.

using SparseNeighbors = std::shared_ptr<NearestNeighbors<SparseVertex>>

Nearest neighbor structure which works over the SpannerGraph.

using DenseGraph = boost::adjacency_list<boost::vecS, boost::vecS, boost::undirectedS, boost::property<vertex_state_t, base::State*, boost::property<boost::vertex_predecessor_t, VertexIndexType, boost::property<boost::vertex_rank_t, VertexIndexType, boost::property<vertex_representative_t, SparseVertex>>>>, boost::property<boost::edge_weight_t, double>>

The underlying roadmap graph.

Any BGL graph representation could be used here. Because we

expect the roadmap to be sparse (m<n^2), an adjacency_list is more appropriate than an adjacency_matrix.

Obviously, a ompl::base::State* vertex property is required.

The incremental connected components algorithm requires vertex_predecessor_t and vertex_rank_t properties. If boost::vecS is not used for vertex storage, then there must also be a boost:vertex_index_t property manually added.

DenseEdges should be undirected and have a weight property.

using DenseVertex = boost::graph_traits<DenseGraph>::vertex_descriptor

A vertex in DenseGraph.

using DenseEdge = boost::graph_traits<DenseGraph>::edge_descriptor

An edge in DenseGraph.

using DenseNeighbors = std::shared_ptr<NearestNeighbors<DenseVertex>>

Nearest neighbor structure which works over the DenseGraph.

Public Functions

SPARS(const base::SpaceInformationPtr &si)

Constructor.

~SPARS() override

Destructor.

void setProblemDefinition(const base::ProblemDefinitionPtr &pdef) override
virtual void getPlannerData(base::PlannerData &data) const override

Get information about the current run of the motion planner. Repeated calls to this function will update data (only additions are made). This is useful to see what changed in the exploration datastructure, between calls to solve(), for example (without calling clear() in between).

void constructRoadmap(const base::PlannerTerminationCondition &ptc)

While the termination condition permits, construct the spanner graph.

void constructRoadmap(const base::PlannerTerminationCondition &ptc, bool stopOnMaxFail)

While the termination condition permits, construct the spanner graph. If stopOnMaxFail is true, the function also terminates when the failure limit set by setMaxFailures() is reached.

virtual base::PlannerStatus solve(const base::PlannerTerminationCondition &ptc) override

Function that can solve the motion planning problem. This function can be called multiple times on the same problem, without calling clear() in between. This allows the planner to continue work for more time on an unsolved problem, for example. Start and goal states from the currently specified ProblemDefinition are cached. This means that between calls to solve(), input states are only added, not removed. When using PRM as a multi-query planner, the input states should be however cleared, without clearing the roadmap itself. This can be done using the clearQuery() function.

virtual void clearQuery() override

Clear the query previously loaded from the ProblemDefinition. Subsequent calls to solve() will reuse the previously computed roadmap, but will clear the set of input states constructed by the previous call to solve(). This enables multi-query functionality for PRM.

virtual void clear() override

Clear all internal datastructures. Planner settings are not affected. Subsequent calls to solve() will ignore all previous work.

template<template<typename T> class NN>
inline void setDenseNeighbors()

Set a different nearest neighbors datastructure for the roadmap graph. This nearest neighbor structure contains only information on the nodes existing in the underlying dense roadmap. This structure is used for near-neighbor queries for the construction of that graph as well as for determining which dense samples the sparse roadmap nodes should represent.

template<template<typename T> class NN>
inline void setSparseNeighbors()

Set a different nearest neighbors datastructure for the spanner graph. This structure is stores only nodes in the roadmap spanner, and is used in the construction of the spanner. It can also be queried to determine which node in the spanner should represent a given state.

inline void setMaxFailures(unsigned int m)

Set the maximum consecutive failures to augment the spanner before termination. In general, if the algorithm fails to add to the spanner for M consecutive iterations, then we can probabilistically estimate how close to attaining the desired properties the SPARS spanner is.

inline void setDenseDeltaFraction(double d)

Set the delta fraction for interface detection. If two nodes in the dense graph are more than a delta fraction of the max. extent apart, then the algorithm cannot consider them to have accurately approximated the location of an interface.

inline void setSparseDeltaFraction(double d)

Set the delta fraction for connection distance on the sparse spanner. This value represents the visibility range of sparse samples. A sparse node represents all dense nodes within a delta fraction of the max. extent if it is also the closest sparse node to that dense node.

inline void setStretchFactor(double t)

Set the roadmap spanner stretch factor. This value represents a multiplicative upper bound on path quality that should be produced by the roadmap spanner. The produced sparse graph with solutions that are less than t times the optimap path length. It does not make sense to make this parameter more than 3.

inline unsigned getMaxFailures() const

Retrieve the maximum consecutive failure limit.

inline double getDenseDeltaFraction() const

Retrieve the dense graph interface support delta fraction.

inline double getSparseDeltaFraction() const

Retrieve the sparse graph visibility range delta fraction.

inline double getStretchFactor() const

Retrieve the spanner’s set stretch factor.

virtual void setup() override

Perform extra configuration steps, if needed. This call will also issue a call to ompl::base::SpaceInformation::setup() if needed. This must be called before solving.

inline const DenseGraph &getDenseGraph() const

Retrieve the underlying dense graph structure. This is built as a PRM* and asymptotically approximates best paths through the space.

inline const SpannerGraph &getRoadmap() const

Retrieve the sparse roadmap structure. This is the structure which answers given queries, and has the desired property of asymptotic near-optimality.

inline unsigned int milestoneCount() const

Returns the number of milestones added to D.

inline unsigned int guardCount() const

Returns the number of guards added to S.

double averageValence() const

Returns the average valence of the spanner graph.

void printDebug(std::ostream &out = std::cout) const

Print debug information about planner.

bool reachedFailureLimit() const

Returns true if we have reached the iteration failures limit, maxFailures_

inline std::string getIterationCount() const
inline std::string getBestCost() const

Protected Functions

DenseVertex addSample(base::State *workState, const base::PlannerTerminationCondition &ptc)

Attempt to add a single sample to the roadmap.

void checkQueryStateInitialization()

Check that the query vertex is initialized (used for internal nearest neighbor searches)

bool sameComponent(SparseVertex m1, SparseVertex m2)

Check that two vertices are in the same connected component.

DenseVertex addMilestone(base::State *state)

Construct a milestone for a given state (state) and store it in the nearest neighbors data structure.

SparseVertex addGuard(base::State *state, GuardType type)

Construct a node with the given state (state) for the spanner and store it in the nn structure.

void connectSparsePoints(SparseVertex v, SparseVertex vp)

Convenience function for creating an edge in the Spanner Roadmap.

void connectDensePoints(DenseVertex v, DenseVertex vp)

Connects points in the dense graph.

bool checkAddCoverage(const base::State *lastState, const std::vector<SparseVertex> &neigh)

Checks the latest dense sample for the coverage property, and adds appropriately.

bool checkAddConnectivity(const base::State *lastState, const std::vector<SparseVertex> &neigh)

Checks the latest dense sample for connectivity, and adds appropriately.

bool checkAddInterface(const std::vector<DenseVertex> &graphNeighborhood, const std::vector<DenseVertex> &visibleNeighborhood, DenseVertex q)

Checks the latest dense sample for bridging an edge-less interface.

bool checkAddPath(DenseVertex q, const std::vector<DenseVertex> &neigh)

Checks for adding an entire dense path to the Sparse Roadmap.

DenseVertex getInterfaceNeighbor(DenseVertex q, SparseVertex rep)

Get the first neighbor of q who has representative rep and is within denseDelta_.

bool addPathToSpanner(const DensePath &dense_path, SparseVertex vp, SparseVertex vpp)

Method for actually adding a dense path to the Roadmap Spanner, S.

void updateRepresentatives(SparseVertex v)

Automatically updates the representatives of all dense samplse within sparseDelta_ of v.

void calculateRepresentative(DenseVertex q)

Calculates the representative for a dense sample.

void addToRepresentatives(DenseVertex q, SparseVertex rep, const std::set<SparseVertex> &oreps)

Adds a dense sample to the appropriate lists of its representative.

void removeFromRepresentatives(DenseVertex q, SparseVertex rep)

Removes the node from its representative’s lists.

void computeVPP(DenseVertex v, DenseVertex vp, std::vector<SparseVertex> &VPPs)

Computes all nodes which qualify as a candidate v” for v and vp.

void computeX(DenseVertex v, DenseVertex vp, DenseVertex vpp, std::vector<SparseVertex> &Xs)

Computes all nodes which qualify as a candidate x for v, v’, and v”.

void resetFailures()

A reset function for resetting the failures count.

void checkForSolution(const base::PlannerTerminationCondition &ptc, base::PathPtr &solution)

Thread that checks for solution

bool haveSolution(const std::vector<DenseVertex> &starts, const std::vector<DenseVertex> &goals, base::PathPtr &solution)

Check if there exists a solution, i.e., there exists a pair of milestones such that the first is in start and the second is in goal, and the two milestones are in the same connected component. If a solution is found, the path is saved.

bool reachedTerminationCriterion() const

Returns true if we have reached the iteration failures limit, maxFailures_ or if a solution was added.

base::PathPtr constructSolution(SparseVertex start, SparseVertex goal) const

Given two milestones from the same connected component, construct a path connecting them and set it as the solution.

void computeDensePath(DenseVertex start, DenseVertex goal, DensePath &path) const

Constructs the dense path between the start and goal vertices (if connected)

void freeMemory()

Free all the memory allocated by the planner.

void getSparseNeighbors(base::State *inState, std::vector<SparseVertex> &graphNeighborhood)

Get all nodes in the sparse graph which are within sparseDelta_ of the given state.

void filterVisibleNeighbors(base::State *inState, const std::vector<SparseVertex> &graphNeighborhood, std::vector<SparseVertex> &visibleNeighborhood) const

Get the visible neighbors.

void getInterfaceNeighborRepresentatives(DenseVertex q, std::set<SparseVertex> &interfaceRepresentatives)

Gets the representatives of all interfaces that q supports.

void getInterfaceNeighborhood(DenseVertex q, std::vector<DenseVertex> &interfaceNeighborhood)

Gets the neighbors of q who help it support an interface.

inline double distanceFunction(const DenseVertex a, const DenseVertex b) const

Compute distance between two milestones (this is simply distance between the states of the milestones)

inline double sparseDistanceFunction(const SparseVertex a, const SparseVertex b) const

Compute distance between two nodes in the sparse roadmap spanner.

base::Cost costHeuristic(SparseVertex u, SparseVertex v) const

Given two vertices, returns a heuristic on the cost of the path connecting them. This method wraps OptimizationObjective::motionCostHeuristic.

Protected Attributes

base::ValidStateSamplerPtr sampler_

Sampler user for generating valid samples in the state space.

DenseNeighbors nn_

Nearest neighbors data structure.

SparseNeighbors snn_

Nearest Neighbors structure for the sparse roadmap.

DenseGraph g_

The dense graph, D.

SpannerGraph s_

The sparse roadmap, S.

std::vector<SparseVertex> startM_

Array of start guards.

std::vector<SparseVertex> goalM_

Array of goal guards.

DenseVertex sparseQueryVertex_

DenseVertex for performing nearest neighbor queries on the SPARSE roadmap.

DenseVertex queryVertex_

Vertex for performing nearest neighbor queries on the DENSE graph.

PathGeometric geomPath_

Geometric Path variable used for smoothing out paths.

boost::property_map<DenseGraph, vertex_state_t>::type stateProperty_

Access to the internal base::state at each DenseVertex.

boost::property_map<SpannerGraph, vertex_state_t>::type sparseStateProperty_

Access to the internal base::State for each SparseVertex of S.

boost::property_map<SpannerGraph, vertex_color_t>::type sparseColorProperty_

Access to draw colors for the SparseVertexs of S, to indicate addition type.

boost::property_map<DenseGraph, vertex_representative_t>::type representativesProperty_

Access to the representatives of the Dense vertices.

boost::property_map<SpannerGraph, vertex_list_t>::type nonInterfaceListsProperty_

Access to all non-interface supporting vertices of the sparse nodes.

boost::property_map<SpannerGraph, vertex_interface_list_t>::type interfaceListsProperty_

Access to the interface-supporting vertice hashes of the sparse nodes.

PathSimplifierPtr psimp_

A path simplifier used to simplify dense paths added to S.

boost::property_map<DenseGraph, boost::edge_weight_t>::type weightProperty_

Access to the weights of each DenseEdge.

boost::disjoint_sets<boost::property_map<SpannerGraph, boost::vertex_rank_t>::type, boost::property_map<SpannerGraph, boost::vertex_predecessor_t>::type> sparseDJSets_

Data structure that maintains the connected components of S.

std::function<const std::vector<DenseVertex>&(const DenseVertex)> connectionStrategy_

Function that returns the milestones to attempt connections with.

unsigned int consecutiveFailures_ = {0u}

A counter for the number of consecutive failed iterations of the algorithm.

double stretchFactor_ = {3.}

The stretch factor in terms of graph spanners for SPARS to check against.

unsigned int maxFailures_ = {1000u}

The maximum number of failures before terminating the algorithm.

bool addedSolution_ = {false}

A flag indicating that a solution has been added during solve()

double denseDeltaFraction_ = {.001}

SPARS parameter for dense graph connection distance as a fraction of max. extent.

double sparseDeltaFraction_ = {.25}

SPARS parameter for Sparse Roadmap connection distance as a fraction of max. extent.

double denseDelta_ = {0.}

SPARS parameter for dense graph connection distance.

double sparseDelta_ = {0.}

SPARS parameter for Sparse Roadmap connection distance.

RNG rng_

Random number generator.

mutable std::mutex graphMutex_

Mutex to guard access to the graphs.

base::OptimizationObjectivePtr opt_

Objective cost function for PRM graph edges.

long unsigned int iterations_ = {0ul}

A counter for the number of iterations of the algorithm.

base::Cost bestCost_ = {std::numeric_limits<double>::quiet_NaN()}

Best cost found so far by algorithm.

struct vertex_color_t

Public Types

using kind = boost::vertex_property_tag
struct vertex_interface_list_t

Public Types

using kind = boost::vertex_property_tag
struct vertex_list_t

Public Types

using kind = boost::vertex_property_tag
struct vertex_representative_t

Public Types

using kind = boost::vertex_property_tag
struct vertex_state_t

Public Types

using kind = boost::vertex_property_tag