DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS > Class Template Reference

This class allows users to solve direct trajectory optimization problems. using different solvers (ipopt/snopt) and different direct methods for trajectory optimization. More...

#include <direct_trajectory_optimization_problem.hpp>

Public Member Functions
	~DirectTrajectoryOptimizationProblem ()
	Destruct the given DTOSolver instance.
void	InitializeSolver (const std::string &problemName, const std::string &outputFile, const std::string &paramFile, const bool &computeJacobian)
	Initialize the DTOSolver problem. More...
void	SetInitialState (const state_vector_t &state)
	Set initial state as hard constraint.
void	SetFinalState (const state_vector_t &state)
	Set final state as hard constraint.
void	SetFinalVelocityZero (const Eigen::VectorXi &vel_indexes)
	Set final velocity zero as hard constraint.
void	SetInitialControl (const control_vector_t &control)
	Set initial control.
void	SetFinalControl (const control_vector_t &control)
	Set final control.
void	SetEvenTimeIncrements (const bool &opt)
	Set even time increments as hard constraint.
void	SetPercentageNodesPerDynamics (Eigen::VectorXd nodeSplit)
	Set percentages of nodes per phase. More...
void	SetGuardFunction (std::vector< double(*)(state_vector_t &)> fcnVector)
	Set pointer to function evaluating guard function. More...
void	Solve (bool warminit=false)
	Solves the DTO problem. More...
void	SetVerbosity (bool verbose)
	Change Verbosity level for solver and method. More...
void	GetDTOSolution (state_vector_array_t &yTraj, control_vector_array_t &uTraj, Eigen::VectorXd &hTraj)
	Get state control and time solution. More...
void	GetDTOSolution (state_vector_array_t &yTraj, control_vector_array_t &uTraj, Eigen::VectorXd &hTraj, std::vector< int > &phaseId)
	Get state control , time and phaseId solution. More...
void	GetFVector (Eigen::VectorXd &fVec)
	Get final value of constraint vector. More...
int	ChangeNumberOfPoints ()
	Change the number of points after construction.
void	SetStateBounds (const state_vector_t &lb, const state_vector_t &ub)
	Set constant state trajectory bounds. More...
void	SetStateBounds (const state_vector_array_t &lb, const state_vector_array_t &ub)
	Set trajectory of state bounds. More...
void	SetControlBounds (const control_vector_t &lb, const control_vector_t &ub)
	Set constant control trajectory bounds. More...
void	SetControlBounds (const control_vector_array_t &lb, const control_vector_array_t &ub)
	Set trajectory of control bounds. More...
void	SetFreeFinalState (int state_number)
	Set final state free of bounds. More...
void	SetTimeIncrementsBounds (const double &h_min, const double &h_max)
	Set bounds for the time increments between nodes. More...
void	SetDircolMethod (const int &dcm)
	SetDircolMethod. More...

Static Public Member Functions
static std::unique_ptr < DirectTrajectoryOptimizationProblem < DIMENSIONS > >	CreateWithMultipleDynamics (std::vector< std::shared_ptr< DynamicsBase< DIMENSIONS > > > dynamics, std::vector< std::shared_ptr< DerivativesBaseDS< DIMENSIONS > > > derivatives, std::shared_ptr< BaseClass::CostFunctionBase< DIMENSIONS > > costFunction, std::vector< std::shared_ptr< BaseClass::ConstraintsBase< DIMENSIONS > > > constraints, Eigen::Vector2d duration, int numberOfNodes, Solvers::solver_t solverType, Methods::method_t methodType, bool solverVerbose, bool methodVerbose, bool feasiblePointOnly=false)
	Create a new DTOSolver instance. More...
static std::unique_ptr < DirectTrajectoryOptimizationProblem < DIMENSIONS > >	Create (std::shared_ptr< DynamicsBase< DIMENSIONS > > dynamics, std::shared_ptr< DerivativesBaseDS< DIMENSIONS > > derivatives, std::shared_ptr< BaseClass::CostFunctionBase< DIMENSIONS > > costFunction, std::shared_ptr< BaseClass::ConstraintsBase< DIMENSIONS > > constraints, Eigen::Vector2d duration, int numberOfNodes, Solvers::solver_t solverType, Methods::method_t methodType, bool solverVerbose, bool methodVerbose, bool feasiblePointOnly=false)
	Create a new DTOSolver instance for single phase problem. More...

Public Attributes
std::shared_ptr < DTOMethodInterface < DIMENSIONS > >	_method
std::shared_ptr < DTOSolverInterface < DIMENSIONS > >	_solver

Detailed Description

template<typename DIMENSIONS>
class DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >

This class allows users to solve direct trajectory optimization problems. using different solvers (ipopt/snopt) and different direct methods for trajectory optimization.

Template Parameters

DIMENSIONS

the class defining the vector types and structures with adequate sizes.

Member Function Documentation

template<typename DIMENSIONS >

static std::unique_ptr<DirectTrajectoryOptimizationProblem<DIMENSIONS> > DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::Create ( std::shared_ptr< DynamicsBase< DIMENSIONS > >  dynamics, std::shared_ptr< DerivativesBaseDS< DIMENSIONS > >  derivatives, std::shared_ptr< BaseClass::CostFunctionBase< DIMENSIONS > >  costFunction, std::shared_ptr< BaseClass::ConstraintsBase< DIMENSIONS > >  constraints, Eigen::Vector2d  duration, int  numberOfNodes, Solvers::solver_t  solverType, Methods::method_t  methodType, bool  solverVerbose, bool  methodVerbose, bool  feasiblePointOnly = false  )

inlinestatic

Create a new DTOSolver instance for single phase problem.

Parameters

[in]	dynamics	The dynamics of your problem.
[in]	derivatives	The derivatives of your problem.
[in]	costFunction	The chosen cost function for the NLP.
[in]	constraints	The chosen vector of constraints for the NLP.
[in]	duration	Time bounds (Tmin, Tmax)
[in]	numberOfNodes	Number of nodes for trajectory discretization.
[in]	solverType	The solver of to employ for the solution of this NLP - see related enum for options.
[in]	methodType	The particular direct method to employ for trajectory optimization - see related enum for options.
[in]	solverVerbose	Show output of solver iterations.
[in]	methodVerbose	Show output of discretization process.
[in]	feasiblePointOnly	The type of optimization.

Returns

A unique pointer to the created solver.

Examples:

acrobot_ipopt_example.cpp, and acrobot_snopt_example.cpp.

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::CreateWithMultipleDynamics().

template<typename DIMENSIONS >

static std::unique_ptr<DirectTrajectoryOptimizationProblem<DIMENSIONS> > DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::CreateWithMultipleDynamics ( std::vector< std::shared_ptr< DynamicsBase< DIMENSIONS > > >  dynamics, std::vector< std::shared_ptr< DerivativesBaseDS< DIMENSIONS > > >  derivatives, std::shared_ptr< BaseClass::CostFunctionBase< DIMENSIONS > >  costFunction, std::vector< std::shared_ptr< BaseClass::ConstraintsBase< DIMENSIONS > > >  constraints, Eigen::Vector2d  duration, int  numberOfNodes, Solvers::solver_t  solverType, Methods::method_t  methodType, bool  solverVerbose, bool  methodVerbose, bool  feasiblePointOnly = false  )

inlinestatic

Create a new DTOSolver instance.

Parameters

[in]	dynamics	Vector of the dynamics of your problem.
[in]	derivatives	Vector of derivatives of your problem.
[in]	costFunction	The chosen cost function for the NLP.
[in]	constraints	The chosen vector of constraints for the NLP.
[in]	duration	Time bounds (Tmin, Tmax).
[in]	numberOfNodes	Number of nodes for trajectory discretization.
[in]	solverType	The solver of to employ for the solution of this NLP - see related enum for options.
[in]	methodType	The particular direct method to employ for trajectory optimization - see related enum for options.
[in]	solverVerbose	Show output of solver iterations.
[in]	methodVerbose	Show output of discretization process.
[in]	feasiblePointOnly	The type of optimization.

Returns

A unique pointer to the created solver.

References DirectTrajectoryOptimization::Methods::DIRECT_TRANSCRIPTION, DirectTrajectoryOptimization::Solvers::IPOPT_SOLVER, DirectTrajectoryOptimization::Methods::MULTIPLE_SHOOTING, and DirectTrajectoryOptimization::Solvers::SNOPT_SOLVER.

Referenced by DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::Create().

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::GetDTOSolution ( state_vector_array_t &  yTraj, control_vector_array_t &  uTraj, Eigen::VectorXd &  hTraj  )

inline

Get state control and time solution.

Parameters

[out]	yTraj	The state trajectory solution
[out]	uTraj	The control trajectory solution
[out]	hTraj	the time increments solution

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_solver.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::GetDTOSolution ( state_vector_array_t &  yTraj, control_vector_array_t &  uTraj, Eigen::VectorXd &  hTraj, std::vector< int > &  phaseId  )

inline

Get state control , time and phaseId solution.

Parameters

[out]	yTraj	The state trajectory solution
[out]	uTraj	The control trajectory solution
[out]	hTraj	the time increments solution
[out]	phaseId	The phase Id solution

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_solver.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::GetFVector ( Eigen::VectorXd &  fVec )

inline

Get final value of constraint vector.

Parameters

[out]

fVec

The vector of cost function and constraints

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_solver.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::InitializeSolver ( const std::string &  problemName, const std::string &  outputFile, const std::string &  paramFile, const bool &  computeJacobian  )

inline

Initialize the DTOSolver problem.

Parameters

[in]	problemName	The name of the problem.
[in]	outputFile	The name of the output file.
[in]	paramFile	The parameters file (non implemented)
[in]	computeJacobian	Flag indicating user is providing Jacobians

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_solver.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetControlBounds ( const control_vector_t &  lb, const control_vector_t &  ub  )

inline

Set constant control trajectory bounds.

Parameters

[in]	lb	control lower bound
[in]	ub	control upper bound

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_method.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetControlBounds ( const control_vector_array_t &  lb, const control_vector_array_t &  ub  )

inline

Set trajectory of control bounds.

Parameters

[in]	lb	control trajectory of lower bounds
[in]	ub	control trajectory of upper bounds

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_method.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetDircolMethod ( const int &  dcm )

inline

SetDircolMethod.

Parameters

[in]

dmc

1 The type of direct collocation method (1-Hermite-Simpson 0-Trapezoidal)

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_method.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetFreeFinalState ( int  state_number )

inline

Set final state free of bounds.

Parameters

[in]

state_number

The index of the state to be free

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_method.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetGuardFunction ( std::vector< double(*)(state_vector_t &)>  fcnVector )

inline

Set pointer to function evaluating guard function.

Parameters

[in]

fcnVector

Pointer to the guard function

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_method.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetPercentageNodesPerDynamics ( Eigen::VectorXd  nodeSplit )

inline

Set percentages of nodes per phase.

Parameters

[in]

nodeSplit

Eigen integer vector indicating percentage of nodes for each phase

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_method.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetStateBounds ( const state_vector_t &  lb, const state_vector_t &  ub  )

inline

Set constant state trajectory bounds.

Parameters

[in]	lb	state lower bound
[in]	ub	state upper bound

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_method.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetStateBounds ( const state_vector_array_t &  lb, const state_vector_array_t &  ub  )

inline

Set trajectory of state bounds.

Parameters

[in]	lb	state trajectory of lower bounds
[in]	ub	state trajectory of upper bounds

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_method.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetTimeIncrementsBounds ( const double &  h_min, const double &  h_max  )

inline

Set bounds for the time increments between nodes.

Parameters

[in]	h_min	positive minimum time increment
[in]	h_max	positive maximum time increment

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_method.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetVerbosity ( bool  verbose )

inline

Change Verbosity level for solver and method.

Parameters

[in]

verbose

Flag indicating verbosity

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_method, and DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_solver.

template<typename DIMENSIONS >

void DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::Solve ( bool  warminit = false )

inline

Solves the DTO problem.

Parameters

[in]

warminit

Flag indicating warmstart

References DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_solver.

Member Data Documentation

template<typename DIMENSIONS >

std::shared_ptr<DTOMethodInterface<DIMENSIONS> > DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_method

The pointer to the DT method

Referenced by DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::ChangeNumberOfPoints(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetControlBounds(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetDircolMethod(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetEvenTimeIncrements(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetFinalControl(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetFinalState(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetFinalVelocityZero(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetFreeFinalState(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetGuardFunction(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetInitialControl(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetInitialState(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetPercentageNodesPerDynamics(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetStateBounds(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetTimeIncrementsBounds(), and DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetVerbosity().

template<typename DIMENSIONS >

std::shared_ptr<DTOSolverInterface<DIMENSIONS> > DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::_solver

The pointer to the Solver

Referenced by DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::GetDTOSolution(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::GetFVector(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::InitializeSolver(), DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::SetVerbosity(), and DirectTrajectoryOptimization::DirectTrajectoryOptimizationProblem< DIMENSIONS >::Solve().

---

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

• /home/depardo/catkin_ws/src/direct_trajectoryoptimization/include/direct_trajectory_optimization_problem.hpp