Public Member Functions | Protected Member Functions | Protected Attributes
OCP Class Reference
Basic Data Structures -- used mainly internally or by the user to call interface functions

Data class for defining optimal control problems. More...

#include <ocp.hpp>

Inheritance diagram for OCP:
Inheritance graph
[legend]

List of all members.

Public Member Functions

returnValue maximizeLagrangeTerm (const Expression &arg)
returnValue maximizeMayerTerm (const Expression &arg)
returnValue minimizeLagrangeTerm (const Expression &arg)
returnValue minimizeMayerTerm (const Expression &arg)
returnValue minimizeMayerTerm (const int &multiObjectiveIdx, const Expression &arg)
 OCP (const double &tStart_=0.0, const double &tEnd_=1.0, const int &N_=20)
 OCP (const double &tStart_, const double &tEnd_, const DVector &_numSteps)
 OCP (const double &tStart_, const Parameter &tEnd_, const int &N_=20)
 OCP (const Grid &grid_)
returnValue subjectTo (const DifferentialEquation &differentialEquation_)
returnValue subjectTo (const ConstraintComponent &component)
returnValue subjectTo (int index_, const ConstraintComponent &component)
returnValue subjectTo (const double lb_, const Expression &arg1, const Expression &arg2, const double ub_)
virtual ~OCP ()
Least Squares terms.

Adds a Least Square term of the form

\begin{equation*} \frac{1}{2} \sum_i \Vert h(t_i,x(t_i),u(t_i),p(t_i),...) - r \Vert^2_{S_{i}} \end{equation*}

Here the sum is over all grid points of the objective grid. The DMatrix $ S $ is assumed to be symmetric and positive (semi-) definite. The Function $ r $ is called reference and can be specified by the user. The function $ h $ is a standard Function.

See also:
Function
Returns:
SUCCESSFUL_RETURN
returnValue minimizeLSQ (const DMatrix &S, const Function &h, const DVector &r)
returnValue minimizeLSQ (const Function &h, const DVector &r)
returnValue minimizeLSQ (const Function &h)
returnValue minimizeLSQ (const MatrixVariablesGrid &S, const Function &h, const VariablesGrid &r)
returnValue minimizeLSQ (const DMatrix &S, const Function &h, const VariablesGrid &r)
returnValue minimizeLSQ (const Function &h, const VariablesGrid &r)
returnValue minimizeLSQ (const MatrixVariablesGrid &S, const Function &h, const char *rFilename)
returnValue minimizeLSQ (const DMatrix &S, const Function &h, const char *rFilename)
returnValue minimizeLSQ (const Function &h, const char *rFilename)
returnValue minimizeLSQ (const DMatrix &S, const Function &h)
returnValue minimizeLSQ (const BMatrix &S, const Function &h)
returnValue minimizeLSQ (const DMatrix &S, const std::string &h)
returnValue minimizeLSQ (const BMatrix &S, const std::string &h)
Least Squares end terms.

Adds an Least Square term that is only evaluated at the end:

\begin{equation*} \frac{1}{2} \Vert( m(T,x(T),u(T),p(T),...) - r )\Vert^2_S \end{equation*}

where $ S $ is a weighting matrix, $ r $ a reference vector and $ T $ the time at the last objective grid point. The function $ m $ is a standard Function.

Returns:
SUCCESSFUL_RETURN
returnValue minimizeLSQEndTerm (const DMatrix &S, const Function &m, const DVector &r)
returnValue minimizeLSQEndTerm (const Function &m, const DVector &r)
returnValue minimizeLSQEndTerm (const DMatrix &S, const Function &m)
returnValue minimizeLSQEndTerm (const DMatrix &S, const std::string &m)
returnValue minimizeLSQEndTerm (const BMatrix &S, const Function &m)
returnValue minimizeLSQEndTerm (const BMatrix &S, const std::string &m)
Set linear terms in the LSQ formulation.
returnValue minimizeLSQLinearTerms (const DVector &Slx, const DVector &Slu)
returnValue minimizeLSQLinearTerms (const BVector &Slx, const BVector &Slu)
returnValue subjectTo (const double lb_, const Expression *arguments, const double ub_)
Add a custom constraint to the OCP formulation.

Adds a constraint of the form

\begin{equation*} \text{lb} <= h_i(t_i, x(t_i), u(t_i), p, ...) <= \text{ub} \end{equation*}

with constant lower and upper bounds.

returnValue subjectTo (const DVector &_lb, const Expression &_expr, const DVector &_ub)
returnValue subjectTo (int _index, const DVector &_lb, const Expression &_expr, const DVector &_ub)
Helper functions.
BooleanType hasObjective () const
BooleanType hasConstraint () const
returnValue getGrid (Grid &grid_) const
returnValue getObjective (Objective &objective_) const
returnValue getObjective (const int &multiObjectiveIdx, Expression **arg) const
returnValue getConstraint (Constraint &constraint_) const
returnValue setObjective (const Objective &objective_)
returnValue setConstraint (const Constraint &constraint_)
returnValue setNumberIntegrationSteps (const uint numSteps)
virtual BooleanType hasEquidistantGrid () const
double getStartTime () const
double getEndTime () const

Protected Member Functions

void setupGrid (double tStart, double tEnd, int N)
void setupGrid (const DVector &times)

Protected Attributes

std::tr1::shared_ptr< Constraint > constraint
std::tr1::shared_ptr< Grid > grid
std::tr1::shared_ptr< Objective > objective

Detailed Description

Data class for defining optimal control problems.

The class OCP is a data class for defining optimal control problems. In the most easiest an optimal control problem can consists of an objective only - i.e. in principle we can set up NLP's as well if no dynamic system is specified. However, in general the objective functional is optimized subject to a dynamic equation and different kind of constraints.

Note that the OCP class is only designed for the user and collects data only. In order to solve an OCP this class must be setup and passed to an appropriate OptimizationAlgorithm.

For setting up an optimal control problem (OCP), we should first specify the time horizon on which the OCP is defined. Note that there are several constructors available which allow to construct an OCP directly with the corresponding time interval. Here, the interval can consist of of given bounds, but in another variant a parameter can be passed in order to allow the setup of optimal control problems for which the end time is optimized, too.

Constraints can be specified with the "subjectTo" syntax. Please note that every parameter, state, control etc which is not fixed via a constraint will be regarded as an optimization variable. In particular, initial value or boundary constraints appear in many OCP formulations and of course all these constraints should all be set explicitly. Moreover, the dynamic equations (model) is regarded as a constraint, too.

Please note that the OCP class only collects the formulation of the problem. If initial values for non-linear problems should be specified, this needs to be done on the algorithm dealing with the OCP. (

See also:
OptimizationAlgorithm)

For advanced users and developers it might be important to know that the class OCP inherits the MultiObjectiveFunctionality which is needed if more than one objective should be specified.

Please check the tutorial examples as well as the class reference below, to learn about the usage of this class in more detail.

Definition at line 89 of file ocp.hpp.

Constructor & Destructor Documentation

BEGIN_NAMESPACE_ACADO OCP::OCP ( const double &  tStart_ = 0.0,
const double &  tEnd_ = 1.0,
const int &  N_ = 20 
)

Default Constructor which can optionally set the time-horizon of the problem.

Parameters:
tStart_start of the time horizon of the OCP
tEnd_end of the time horizon of the OCP
N_number of discretization intervals

Definition at line 45 of file ocp.cpp.

OCP::OCP ( const double &  tStart_,
const double &  tEnd_,
const DVector &  _numSteps 
)

Constructor which can set a non equidistant time-horizon of the problem.

Parameters:
tStart_start of the time horizon of the OCP
tEnd_end of the time horizon of the OCP
_numStepsnumber of integration steps in each discretization interval

Definition at line 56 of file ocp.cpp.

OCP::OCP ( const double &  tStart_,
const Parameter &  tEnd_,
const int &  N_ = 20 
)

Constructor that takes a parametric version of the time horizon. This contructor should be used if the end time should be optimized, too.

Parameters:
tStart_start of the time horizon of the OCP
tEnd_end of the time horizon of the OCP
N_number of discretization intervals

Definition at line 91 of file ocp.cpp.

OCP::OCP ( const Grid &  grid_)

Constructor that takes the time horizon in form of a Grid.

Parameters:
grid_discretization grid

Definition at line 80 of file ocp.cpp.

OCP::~OCP ( ) [virtual]

Destructor (deletes everything).

Definition at line 101 of file ocp.cpp.

Member Function Documentation

returnValue OCP::getConstraint ( Constraint &  constraint_) const

Definition at line 331 of file ocp.cpp.

double OCP::getEndTime ( ) const

Definition at line 362 of file ocp.cpp.

returnValue OCP::getGrid ( Grid &  grid_) const

Definition at line 329 of file ocp.cpp.

returnValue OCP::getObjective ( Objective &  objective_) const

Definition at line 330 of file ocp.cpp.

returnValue OCP::getObjective ( const int &  multiObjectiveIdx,
Expression **  arg 
) const

Reimplemented from MultiObjectiveFunctionality.

Definition at line 355 of file ocp.cpp.

double OCP::getStartTime ( ) const

Definition at line 361 of file ocp.cpp.

BooleanType OCP::hasConstraint ( ) const

Definition at line 322 of file ocp.cpp.

BooleanType OCP::hasEquidistantGrid ( ) const [virtual]

Returns whether the ocp grid is equidistant.

Returns:
true iff the OCP grid is equidistant, false otherwise.

Definition at line 364 of file ocp.cpp.

BooleanType OCP::hasObjective ( ) const

Definition at line 316 of file ocp.cpp.

returnValue OCP::maximizeLagrangeTerm ( const Expression &  arg)

Adds an expression as a the Lagrange term to be maximized.

Returns:
SUCCESSFUL_RETURN

Definition at line 241 of file ocp.cpp.

returnValue OCP::maximizeMayerTerm ( const Expression &  arg)

Adds an expression as a the Mayer term to be maximized. Note that this function is introduced for convenience only. A call of the form maximizeMayerTerm( arg ) is equivalent to calling minimizeMayerTerm( -arg ).

Returns:
SUCCESSFUL_RETURN

Definition at line 239 of file ocp.cpp.

returnValue OCP::minimizeLagrangeTerm ( const Expression &  arg)

Adds an expression as a the Lagrange term to be minimized.

Returns:
SUCCESSFUL_RETURN

Definition at line 240 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const DMatrix &  S,
const Function &  h,
const DVector &  r 
)
Parameters:
Sa weighting matrix
hthe LSQ-Function
rthe reference

Definition at line 244 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const Function &  h,
const DVector &  r 
)
Parameters:
hthe LSQ-Function
rthe reference

Definition at line 252 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const Function &  h)
Parameters:
hthe LSQ-Function

Definition at line 260 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const MatrixVariablesGrid &  S,
const Function &  h,
const VariablesGrid &  r 
)
Parameters:
Sa weighting matrix
hthe LSQ-Function
rthe reference

Definition at line 272 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const DMatrix &  S,
const Function &  h,
const VariablesGrid &  r 
)
Parameters:
Sa weighting matrix
hthe LSQ-Function
rthe reference

Definition at line 279 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const Function &  h,
const VariablesGrid &  r 
)
Parameters:
hthe LSQ-Function
rthe reference

Definition at line 290 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const MatrixVariablesGrid &  S,
const Function &  h,
const char *  rFilename 
)
Parameters:
Sa weighting matrix
hthe LSQ-Function
rFilenamefilename where the reference is stored

Definition at line 110 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const DMatrix &  S,
const Function &  h,
const char *  rFilename 
)
Parameters:
Sa weighting matrix
hthe LSQ-Function
rFilenamefilename where the reference is stored

Definition at line 125 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const Function &  h,
const char *  rFilename 
)
Parameters:
hthe LSQ-Function
rFilenamefilename where the reference is stored

Definition at line 139 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const DMatrix &  S,
const Function &  h 
)
Note:
Applicable only for automatic code generation.
Parameters:
Sa weighting matrix
hthe LSQ-Function

Definition at line 371 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const BMatrix &  S,
const Function &  h 
)

Pass the sparsity pattern of the weighting matrix to the code generator.

Note:
Applicable only for automatic code generation.
Parameters:
Sa weighting matrix
hthe LSQ-Function

Definition at line 381 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const DMatrix &  S,
const std::string &  h 
)
Note:
Applicable only for automatic code generation.
Parameters:
Sa weighting matrix
hthe externally defined LSQ-Function

Definition at line 391 of file ocp.cpp.

returnValue OCP::minimizeLSQ ( const BMatrix &  S,
const std::string &  h 
)

Pass the sparsity pattern of the weighting matrix to the code generator.

Note:
Applicable only for automatic code generation.
Parameters:
Sa weighting matrix
hthe externally defined LSQ-Function

Definition at line 401 of file ocp.cpp.

returnValue OCP::minimizeLSQEndTerm ( const DMatrix &  S,
const Function &  m,
const DVector &  r 
)
Parameters:
Sa weighting matrix
mthe LSQ-Function
rthe reference

Definition at line 297 of file ocp.cpp.

returnValue OCP::minimizeLSQEndTerm ( const Function &  m,
const DVector &  r 
)
Parameters:
mthe LSQ-Function
rthe reference

Definition at line 307 of file ocp.cpp.

returnValue OCP::minimizeLSQEndTerm ( const DMatrix &  S,
const Function &  m 
)
Note:
Applicable only for automatic code generation.
Parameters:
Sa weighting matrix
mthe LSQ-Function

Definition at line 376 of file ocp.cpp.

returnValue OCP::minimizeLSQEndTerm ( const DMatrix &  S,
const std::string &  m 
)
Note:
Applicable only for automatic code generation.
Parameters:
Sa weighting matrix
mthe externally defined LSQ-Function

Definition at line 396 of file ocp.cpp.

returnValue OCP::minimizeLSQEndTerm ( const BMatrix &  S,
const Function &  m 
)

Pass the sparsity pattern of the weighting matrix to the code generator.

Note:
Applicable only for automatic code generation.
Parameters:
Sa weighting matrix
mthe LSQ-Function

Definition at line 386 of file ocp.cpp.

returnValue OCP::minimizeLSQEndTerm ( const BMatrix &  S,
const std::string &  m 
)

Pass the sparsity pattern of the weighting matrix to the code generator.

Note:
Applicable only for automatic code generation.
Parameters:
Sa weighting matrix
mthe externally defined LSQ-Function

Definition at line 406 of file ocp.cpp.

returnValue OCP::minimizeLSQLinearTerms ( const DVector &  Slx,
const DVector &  Slu 
)

Applicable only for automatic code generation.

Note:
Experimental.
Parameters:
Slxa weighting vector for differential states.
Slua weighting vector for controls.

Definition at line 411 of file ocp.cpp.

returnValue OCP::minimizeLSQLinearTerms ( const BVector &  Slx,
const BVector &  Slu 
)

Applicable only for automatic code generation.

Note:
Experimental.
Parameters:
Slxa weighting vector for differential states.
Slua weighting vector for controls.

Definition at line 416 of file ocp.cpp.

returnValue OCP::minimizeMayerTerm ( const Expression &  arg)

Adds an expression as a the Mayer term to be minimized.

Returns:
SUCCESSFUL_RETURN

Definition at line 238 of file ocp.cpp.

returnValue OCP::minimizeMayerTerm ( const int &  multiObjectiveIdx,
const Expression &  arg 
)

Adds an expression as a the Mayer term to be minimized. In this version the number of the objective can be specified as well. This functionality is needed for multi-objective optimal control problems.

Returns:
SUCCESSFUL_RETURN

Reimplemented from MultiObjectiveFunctionality.

Definition at line 104 of file ocp.cpp.

returnValue OCP::setConstraint ( const Constraint &  constraint_)

Definition at line 339 of file ocp.cpp.

returnValue OCP::setNumberIntegrationSteps ( const uint  numSteps)

Definition at line 346 of file ocp.cpp.

returnValue OCP::setObjective ( const Objective &  objective_)

Definition at line 334 of file ocp.cpp.

void OCP::setupGrid ( double  tStart,
double  tEnd,
int  N 
) [protected]

Definition at line 423 of file ocp.cpp.

void OCP::setupGrid ( const DVector &  times) [protected]

Definition at line 432 of file ocp.cpp.

returnValue OCP::subjectTo ( const DifferentialEquation &  differentialEquation_)

Adds an differential equation (as a continuous equality constraint).

Parameters:
differentialEquation_the differential equation to be added
n_the number of control intervals

Returns:
SUCCESSFUL_RETURN

Definition at line 153 of file ocp.cpp.

returnValue OCP::subjectTo ( const ConstraintComponent &  component)

Adds a (continuous) constraint.

Returns:
SUCCESSFUL_RETURN
RET_INFEASIBLE_CONSTRAINT
Adds a discrete, point, constraint.
SUCCESSFUL_RETURN
RET_INFEASIBLE_CONSTRAINT

Definition at line 161 of file ocp.cpp.

returnValue OCP::subjectTo ( int  index_,
const ConstraintComponent &  component 
)

Definition at line 170 of file ocp.cpp.

returnValue OCP::subjectTo ( const double  lb_,
const Expression &  arg1,
const Expression &  arg2,
const double  ub_ 
)

Add a coupled boundary constraint.

Coupled boundary constraints of general form

\begin{equation*} \text{lb} \leq h_1( t_0,x(t_0),u(t_0),p,... ) + h_2( t_e,x(t_e),u(t_e),p,... ) \leq \text{ub} \end{equation*}

where $ t_0 $ is the first and $ t_e $ the last time point in the grid. Adds a constraint of the form 

lb <= arg1(0) + arg2(T) <= ub

with constant lower and upper bounds.

Returns:
SUCCESSFUL_RETURN
RET_INFEASIBLE_CONSTRAINT

Definition at line 194 of file ocp.cpp.

returnValue OCP::subjectTo ( const double  lb_,
const Expression *  arguments,
const double  ub_ 
)

Add a custom constraint.

Adds a constraint of the form

\begin{equation*} \text{lb} <= \sum_i h_i(t_i, x(t_i), u(t_i), p, ...) <= \text{ub} \end{equation*}

with constant lower and upper bounds.

Returns:
SUCCESSFUL_RETURN
RET_INFEASIBLE_CONSTRAINT

Definition at line 201 of file ocp.cpp.

returnValue OCP::subjectTo ( const DVector &  _lb,
const Expression &  _expr,
const DVector &  _ub 
)

Definition at line 206 of file ocp.cpp.

returnValue OCP::subjectTo ( int  _index,
const DVector &  _lb,
const Expression &  _expr,
const DVector &  _ub 
)

Definition at line 214 of file ocp.cpp.

Member Data Documentation

std::tr1::shared_ptr<Constraint> OCP::constraint [protected]

The Constraints.

Definition at line 400 of file ocp.hpp.

std::tr1::shared_ptr<Grid> OCP::grid [protected]

Common discretization grid.

Definition at line 396 of file ocp.hpp.

std::tr1::shared_ptr<Objective> OCP::objective [protected]

The Objective.

Definition at line 398 of file ocp.hpp.

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

acado
Author(s): Milan Vukov, Rien Quirynen
autogenerated on Thu Aug 27 2015 12:01:39