aligator  0.6.1
A primal-dual augmented Lagrangian-type solver for nonlinear trajectory optimization.
Loading...
Searching...
No Matches
Public Types | Public Member Functions | Static Public Member Functions | Public Attributes | Protected Attributes | List of all members
aligator::SolverFDDPTpl< Scalar > Struct Template Reference

The feasible DDP (FDDP) algorithm, from Mastalli et al. (2020). More...

#include <aligator/solvers/fddp/solver-fddp.hpp>

Collaboration diagram for aligator::SolverFDDPTpl< Scalar >:
[legend]

Public Types

using Problem = TrajOptProblemTpl<Scalar>
 
using StageModel = StageModelTpl<Scalar>
 
using StageData = StageDataTpl<Scalar>
 
using ProblemData = TrajOptDataTpl<Scalar>
 
using Results = ResultsFDDPTpl<Scalar>
 
using Workspace = WorkspaceFDDPTpl<Scalar>
 
using Manifold = ManifoldAbstractTpl<Scalar>
 
using VParams = ValueFunctionTpl<Scalar>
 
using QParams = QFunctionTpl<Scalar>
 
using CostData = CostDataAbstractTpl<Scalar>
 
using ExpModel = ExplicitDynamicsModelTpl<Scalar>
 
using ExplicitDynamicsData = ExplicitDynamicsDataTpl<Scalar>
 
using CallbackPtr = shared_ptr<CallbackBaseTpl<Scalar>>
 
using CallbackMap = std::unordered_map<std::string, CallbackPtr>
 

Public Member Functions

 ALIGATOR_DYNAMIC_TYPEDEFS (Scalar)
 
void setNumThreads (const std::size_t num_threads)
 
std::size_t getNumThreads () const
 
 SolverFDDPTpl (const Scalar tol=1e-6, VerboseLevel verbose=VerboseLevel::QUIET, const Scalar reg_init=1e-9, const std::size_t max_iters=200)
 
ALIGATOR_DEPRECATED const Results & getResults () const
 Get the solver results.
 
ALIGATOR_DEPRECATED const Workspace & getWorkspace () const
 Get a const reference to the solver's workspace.
 
void setup (const Problem &problem)
 Allocate workspace and results structs.
 
void updateExpectedImprovement (Workspace &workspace, Results &results) const
 Pre-compute parts of the directional derivatives – this is done before linesearch.
 
void expectedImprovement (Workspace &workspace, Scalar &d1, Scalar &d2) const
 Finish computing the directional derivatives – this is done within linesearch.
 

Static Public Member Functions

static Scalar forwardPass (const Problem &problem, const Results &results, Workspace &workspace, const Scalar alpha)
 Perform a nonlinear rollout, keeping an infeasibility gap.
 

Public Attributes

Scalar target_tol_
 
Scalar th_grad_ = 1e-12
 
Scalar th_step_dec_ = 0.5
 
Scalar th_step_inc_ = 0.01
 
Linesearch< Scalar >::Options ls_params
 
VerboseLevel verbose_
 
std::size_t max_iters
 Maximum number of iterations for the solver.
 
bool force_initial_condition_
 
Logger logger {}
 
Results results_
 
Workspace workspace_
 
Regularization parameters
Scalar reg_init
 
Scalar preg_ = reg_init
 
Scalar reg_min_ = 1e-9
 
Scalar reg_max_ = 1e9
 
Scalar reg_dec_factor_ = 0.1
 Regularization decrease factor.
 
Scalar reg_inc_factor_ = 10.
 Regularization increase factor.
 

Protected Attributes

std::size_t num_threads_
 Number of threads to use when evaluating the problem or its derivatives.
 
CallbackMap callbacks_
 Callbacks.
 

Detailed Description

template<typename Scalar>
struct aligator::SolverFDDPTpl< Scalar >

The feasible DDP (FDDP) algorithm, from Mastalli et al. (2020).

The implementation very similar to Crocoddyl's SolverFDDP.

Definition at line 31 of file solver-fddp.hpp.

Member Typedef Documentation

◆ Problem

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::Problem = TrajOptProblemTpl<Scalar>

Definition at line 33 of file solver-fddp.hpp.

◆ StageModel

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::StageModel = StageModelTpl<Scalar>

Definition at line 34 of file solver-fddp.hpp.

◆ StageData

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::StageData = StageDataTpl<Scalar>

Definition at line 35 of file solver-fddp.hpp.

◆ ProblemData

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::ProblemData = TrajOptDataTpl<Scalar>

Definition at line 36 of file solver-fddp.hpp.

◆ Results

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::Results = ResultsFDDPTpl<Scalar>

Definition at line 37 of file solver-fddp.hpp.

◆ Workspace

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::Workspace = WorkspaceFDDPTpl<Scalar>

Definition at line 38 of file solver-fddp.hpp.

◆ Manifold

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::Manifold = ManifoldAbstractTpl<Scalar>

Definition at line 39 of file solver-fddp.hpp.

◆ VParams

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::VParams = ValueFunctionTpl<Scalar>

Definition at line 40 of file solver-fddp.hpp.

◆ QParams

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::QParams = QFunctionTpl<Scalar>

Definition at line 41 of file solver-fddp.hpp.

◆ CostData

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::CostData = CostDataAbstractTpl<Scalar>

Definition at line 42 of file solver-fddp.hpp.

◆ ExpModel

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::ExpModel = ExplicitDynamicsModelTpl<Scalar>

Definition at line 43 of file solver-fddp.hpp.

◆ ExplicitDynamicsData

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::ExplicitDynamicsData = ExplicitDynamicsDataTpl<Scalar>

Definition at line 44 of file solver-fddp.hpp.

◆ CallbackPtr

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::CallbackPtr = shared_ptr<CallbackBaseTpl<Scalar>>

Definition at line 45 of file solver-fddp.hpp.

◆ CallbackMap

template<typename Scalar >
using aligator::SolverFDDPTpl< Scalar >::CallbackMap = std::unordered_map<std::string, CallbackPtr>

Definition at line 46 of file solver-fddp.hpp.

Constructor & Destructor Documentation

◆ SolverFDDPTpl()

template<typename Scalar >
aligator::SolverFDDPTpl< Scalar >::SolverFDDPTpl ( const Scalar tol = 1e-6,
VerboseLevel verbose = VerboseLevel::QUIET,
const Scalar reg_init = 1e-9,
const std::size_t max_iters = 200 )

Member Function Documentation

◆ ALIGATOR_DYNAMIC_TYPEDEFS()

template<typename Scalar >
aligator::SolverFDDPTpl< Scalar >::ALIGATOR_DYNAMIC_TYPEDEFS ( Scalar )

◆ setNumThreads()

template<typename Scalar >
void aligator::SolverFDDPTpl< Scalar >::setNumThreads ( const std::size_t num_threads)
inline

Definition at line 78 of file solver-fddp.hpp.

◆ getNumThreads()

template<typename Scalar >
std::size_t aligator::SolverFDDPTpl< Scalar >::getNumThreads ( ) const
inline

Definition at line 82 of file solver-fddp.hpp.

◆ getResults()

template<typename Scalar >
ALIGATOR_DEPRECATED const Results & aligator::SolverFDDPTpl< Scalar >::getResults ( ) const
inline

Get the solver results.

Definition at line 100 of file solver-fddp.hpp.

◆ getWorkspace()

template<typename Scalar >
ALIGATOR_DEPRECATED const Workspace & aligator::SolverFDDPTpl< Scalar >::getWorkspace ( ) const
inline

Get a const reference to the solver's workspace.

Definition at line 102 of file solver-fddp.hpp.

◆ setup()

template<typename Scalar >
void aligator::SolverFDDPTpl< Scalar >::setup ( const Problem & problem)

Allocate workspace and results structs.

◆ forwardPass()

template<typename Scalar >
static Scalar aligator::SolverFDDPTpl< Scalar >::forwardPass ( const Problem & problem,
const Results & results,
Workspace & workspace,
const Scalar alpha )
static

Perform a nonlinear rollout, keeping an infeasibility gap.

Perform a nonlinear rollout using the computed sensitivity gains from the backward pass, while keeping the dynamical feasibility gaps open proportionally to the step-size alpha.

Parameters
[in]problem
[in]results
[out]workspace
[in]alphastep-size.

◆ updateExpectedImprovement()

template<typename Scalar >
void aligator::SolverFDDPTpl< Scalar >::updateExpectedImprovement ( Workspace & workspace,
Results & results ) const

Pre-compute parts of the directional derivatives – this is done before linesearch.

Inspired from Crocoddyl's own function, crocoddyl::SolverFDDP::updateExpectedImprovement

◆ expectedImprovement()

template<typename Scalar >
void aligator::SolverFDDPTpl< Scalar >::expectedImprovement ( Workspace & workspace,
Scalar & d1,
Scalar & d2 ) const

Finish computing the directional derivatives – this is done within linesearch.

Inspired from Crocoddyl's own function, crocoddyl::SolverFDDP::expectedImprovement

Member Data Documentation

◆ target_tol_

template<typename Scalar >
Scalar aligator::SolverFDDPTpl< Scalar >::target_tol_

Definition at line 48 of file solver-fddp.hpp.

◆ reg_init

template<typename Scalar >
Scalar aligator::SolverFDDPTpl< Scalar >::reg_init

Definition at line 52 of file solver-fddp.hpp.

◆ preg_

template<typename Scalar >
Scalar aligator::SolverFDDPTpl< Scalar >::preg_ = reg_init

Definition at line 53 of file solver-fddp.hpp.

◆ reg_min_

template<typename Scalar >
Scalar aligator::SolverFDDPTpl< Scalar >::reg_min_ = 1e-9

Definition at line 54 of file solver-fddp.hpp.

◆ reg_max_

template<typename Scalar >
Scalar aligator::SolverFDDPTpl< Scalar >::reg_max_ = 1e9

Definition at line 55 of file solver-fddp.hpp.

◆ reg_dec_factor_

template<typename Scalar >
Scalar aligator::SolverFDDPTpl< Scalar >::reg_dec_factor_ = 0.1

Regularization decrease factor.

Definition at line 57 of file solver-fddp.hpp.

◆ reg_inc_factor_

template<typename Scalar >
Scalar aligator::SolverFDDPTpl< Scalar >::reg_inc_factor_ = 10.

Regularization increase factor.

Definition at line 59 of file solver-fddp.hpp.

◆ th_grad_

template<typename Scalar >
Scalar aligator::SolverFDDPTpl< Scalar >::th_grad_ = 1e-12

Definition at line 62 of file solver-fddp.hpp.

◆ th_step_dec_

template<typename Scalar >
Scalar aligator::SolverFDDPTpl< Scalar >::th_step_dec_ = 0.5

Definition at line 63 of file solver-fddp.hpp.

◆ th_step_inc_

template<typename Scalar >
Scalar aligator::SolverFDDPTpl< Scalar >::th_step_inc_ = 0.01

Definition at line 64 of file solver-fddp.hpp.

◆ ls_params

template<typename Scalar >
Linesearch<Scalar>::Options aligator::SolverFDDPTpl< Scalar >::ls_params

Definition at line 66 of file solver-fddp.hpp.

◆ verbose_

template<typename Scalar >
VerboseLevel aligator::SolverFDDPTpl< Scalar >::verbose_

Definition at line 68 of file solver-fddp.hpp.

◆ max_iters

template<typename Scalar >
std::size_t aligator::SolverFDDPTpl< Scalar >::max_iters

Maximum number of iterations for the solver.

Definition at line 70 of file solver-fddp.hpp.

◆ force_initial_condition_

template<typename Scalar >
bool aligator::SolverFDDPTpl< Scalar >::force_initial_condition_

Crocoddyl's FDDP implementation forces the initial state in linesearch to satisfy the initial condition. This flag switches that behaviour on or off.

Definition at line 74 of file solver-fddp.hpp.

◆ logger

template<typename Scalar >
Logger aligator::SolverFDDPTpl< Scalar >::logger {}

Definition at line 76 of file solver-fddp.hpp.

◆ num_threads_

template<typename Scalar >
std::size_t aligator::SolverFDDPTpl< Scalar >::num_threads_
protected

Number of threads to use when evaluating the problem or its derivatives.

Definition at line 86 of file solver-fddp.hpp.

◆ callbacks_

template<typename Scalar >
CallbackMap aligator::SolverFDDPTpl< Scalar >::callbacks_
protected

Callbacks.

Definition at line 88 of file solver-fddp.hpp.

◆ results_

template<typename Scalar >
Results aligator::SolverFDDPTpl< Scalar >::results_

Definition at line 91 of file solver-fddp.hpp.

◆ workspace_

template<typename Scalar >
Workspace aligator::SolverFDDPTpl< Scalar >::workspace_

Definition at line 92 of file solver-fddp.hpp.

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