Abstract C++ wrapper of generated Acados solver C-code. More...

#include <acados_solver.hpp>

Collaboration diagram for acados::AcadosSolver:

Classes
class	Dimensions
	Container for the (fixed) dimensions of the imported Acados OCP. More...

Public Member Functions
	AcadosSolver ()
	Constructor of the AcadosSolver object.

virtual	~AcadosSolver ()
	Destructor of the AcadosSolver object.

int	init (unsigned int N, double Ts)
	Initialize the solver and set the (constant) sampling intervals.

int	reset ()
	Resets the Acados solver, including the internal QP solver and the RTI phase cache.

int	solve ()
	Solve the non-linear optimization problem given the provided initial state, constraints, etc.

int	solve_rti (RtiStage rti_phase)
	Solve the non-linear (RTI) optimization problem given the provided initial state, constraints, etc.

int	simulate (double dt, ValueVector &x0, ValueVector &u0, ValueVector &p, ValueVector &x_next, ValueVector &z)
	Simulate the next state given the current state, control inputs and runtime parameters.

int	simulate (double dt, ValueMap const &x0_map, ValueMap const &u0_map, ValueMap const &p_map, ValueMap &x_next_map, ValueMap &z_map)
	Simulate the next state given the current state, control inputs and runtime parameters.

int	simulate (ValueMap const &x0_map, ValueMap const &u0_map, ValueMap const &p_map, ValueMap &x_next_map, ValueMap &z_map)
	Simulate the next state given the current state, control inputs and runtime parameters.

int	set_initial_state_values (ValueVector &x_0)
	Set the initial state values (i.e., add constraints on initial state)

int	set_initial_state_values (ValueMap const &x_0_map)
	Set the initial state values (i.e., add constraints on initial state) from a map of key/ValueVector pairs.

int	set_state_bounds (unsigned int stage, IndexVector &idxbx, ValueVector &lbx, ValueVector &ubx)
	Set (differential) state bounds at a given stage.

int	set_state_bounds (IndexVector &idxbx, ValueVector &lbx, ValueVector &ubx)
	Set (differential) state bounds for all stages.

int	set_control_bounds (unsigned int stage, IndexVector &idxbu, ValueVector &lbu, ValueVector &ubu)
	Set control bounds at a given stage.

int	set_control_bounds (IndexVector &idxbu, ValueVector &lbu, ValueVector &ubu)
	Set control bounds for all stages.

int	set_runtime_parameters (unsigned int stage, ValueVector &p_i)
	Set the runtime parameters for one stage from ordered value vector.

int	set_runtime_parameters (unsigned int stage, ValueMap const &p_i_map)
	Set the runtime parameters for one stage from a key/values map.

int	set_runtime_parameters (ValueVector &p_i)
	Set the runtime parameters for all stages at once from ordered value vector.

int	set_runtime_parameters (ValueMap const &p_i_map)
	Set the runtime parameters for all stages at once from a key/values map.

int	initialize_state_values (unsigned int stage, ValueVector &x_i)
	Initialize the (differential) state values for a stage from an ordered value vector.

int	initialize_state_values (unsigned int stage, ValueMap const &x_i_map)
	Initialize the (differential) state values for a stage from a key/values map.

int	initialize_state_values (ValueVector &x_i)
	Initialize the (differential) state values for ALL stages at once from an ordered value vector.

int	initialize_state_values (ValueMap const &x_i_map)
	Initialize the (differential) state values for ALL stages at once from a key/values map.

int	initialize_control_values (unsigned int stage, ValueVector &u_i)
	Initialize the control variable values for one stage from an ordered value vector.

int	initialize_control_values (unsigned int stage, ValueMap const &u_i_map)
	Initialize the control variable values for one stage from a key/values map.

int	initialize_control_values (ValueVector &u_i)
	Initialize the control variable values for ALL stages at once from an ordered value vector.

int	initialize_control_values (ValueMap const &u_i_map)
	Initialize the control variable values for ALL stages at once from a key/values map.

ValueVector	get_state_values (unsigned int stage)
	Retrieve the differential state variables at a given stage.

ValueMap	get_state_values_as_map (unsigned int stage)
	Retrieve the differential state variables at a given stage and package them as a ValueMap.

ValueVector	get_algebraic_state_values (unsigned int stage)
	Retrieve the algebraic state variables at a given stage.

ValueMap	get_algebraic_state_values_as_map (unsigned int stage)
	Retrieve the algebraic state variables at a given stage and package them as a ValueMap.

ValueVector	get_control_values (unsigned int stage)
	Retrieve the control variables at a given stage.

ValueMap	get_control_values_as_map (unsigned int stage)
	Retrieve the control variables at a given stage and package them as a ValueMap.

ValueVector	get_parameter_values (unsigned int stage)
	Retrieve the parameters at a given stage.

ValueMap	get_parameter_values_as_map (unsigned int stage)
	Retrieve the parameters at a given stage and package them as a ValueMap.

const IndexMap &	x_index_map () const
	Returns a constant reference to the Key/IndexVector map of differential state variables internally stored in `AcadosSolver::_x_index_map`.

const IndexMap &	z_index_map () const
	Returns a constant reference to the Key/IndexVector map of algebraic state variables internally stored in `AcadosSolver::_z_index_map`.

const IndexMap &	p_index_map () const
	Returns a constant reference to the Key/IndexVector map of runtime parameters internally stored in `AcadosSolver::_p_index_map`.

const IndexMap &	u_index_map () const
	Returns a constant reference to the Key/IndexVector map of control variables internally stored in `AcadosSolver::_u_index_map`.

const Dimensions &	dims () const
	Returns the (fixed) dimensions of the OCP problem stored in `AcadosSolver::_dims`.

unsigned int	nx () const
	Returns the dimensions of the differential state vector.

unsigned int	nz () const
	Returns the dimensions of the algebraic state vector.

unsigned int	np () const
	Returns the number of runtime parameters.

unsigned int	nu () const
	Returns the number of control variables.

unsigned int	N () const
	Returns the number of shooting nodes.

double	Ts () const
	Returns the (constant) sampling interval Ts (in seconds).

std::vector< double >	sampling_intervals () const
	Returns the vector of sampling interval (in seconds).

virtual ocp_nlp_in *	get_nlp_in () const =0
	Legacy C-function to retrieve the C-interface `ocp_nlp_in` property of the solver.

virtual ocp_nlp_out *	get_nlp_out () const =0
	Legacy C-function to retrieve the C-interface `ocp_nlp_out` property of the solver.

virtual ocp_nlp_out *	get_sens_out () const =0
	Legacy C-function to retrieve the C-interface `ocp_nlp_out` property of the solver.

virtual ocp_nlp_solver *	get_nlp_solver () const =0
	Legacy C-function to retrieve the C-interface `ocp_nlp_solver` property of the solver.

virtual ocp_nlp_config *	get_nlp_config () const =0
	Legacy C-function to retrieve the C-interface `ocp_nlp_config` property of the solver.

virtual void *	get_nlp_opts () const =0
	Legacy C-function to retrieve the C-interface `get_nlp_opts` property of the solver.

virtual ocp_nlp_dims *	get_nlp_dims () const =0
	Legacy C-function to retrieve the C-interface `ocp_nlp_dims` property of the solver that contains the dimensions of the problem.

virtual ocp_nlp_plan_t *	get_nlp_plan () const =0
	Legacy C-function to retrieve the C-interface `ocp_nlp_plan` property of the solver.

virtual unsigned int	get_nlp_np () const =0
	Legacy C-function to retrieve the number of runtime parameters from the Acados capsule.

Static Public Member Functions
static bool	is_map_size_consistent (IndexMap const &index_map, unsigned int expected_total_indexes)
	Check if the total number of indexes contained in the provided index map is equal to an expected number.

static bool	is_map_size_consistent (ValueMap const &values_map, unsigned int expected_total_indexes)
	Check if the total number of values contained in the provided value map is equal to an expected number.

static bool	is_values_map_complete (IndexMap const &index_map, ValueMap const &values_map)
	Check if for each key-index pair contained in the index map there is a value provided by the value map.

static void	fill_vector_from_map (IndexMap const &index_map, ValueMap const &values_map, unsigned int vector_size, ValueVector &values)
	Fill (and resize if necessary) a vector with the values contained in the value map.

static ValueVector	value_vector_from_map (IndexMap const &index_map, ValueMap const &values_map, unsigned int vector_size)
	Returns a vector containing the values provided through the value map.

static ValueMap	create_map_from_values (IndexMap const &index_map, ValueVector const &values)
	Returns a vector ValueMap object containing the values of the value vector.

static void	fill_map_from_values (IndexMap const &index_map, ValueVector const &values, ValueMap &value_map)
	Fill a ValueMap object with the values of the value vector.

Protected Member Functions
virtual int	create_index_maps ()=0
	Reinitialize the maps used to set or retrieve value using string variable names.

int	free_memory ()
	Free all the allocated memory.

virtual int	internal_create_capsule ()=0
	Create the Acados solver's capsule used by the Acados C-interface.

virtual int	internal_create_with_discretization (int n_time_steps, double *new_time_steps)=0
	Initialize Acados solver and (optionally) redefine the shooting nodes.

virtual int	internal_reset (int reset_qp_solver_mem=1)=0
	Reset the solver memory and (opt.) the internal QP solver.

virtual int	internal_free ()=0
	Free the memory allocated for the solver.

virtual int	internal_free_capsule ()=0
	Free the memory allocated for the capsule.

virtual int	internal_update_qp_solver_cond_N (int qp_solver_cond_N)=0
	Update the QP solver `qp_solver_cond_N` setting.

virtual int	internal_update_params (unsigned int stage, double *value, int np)=0
	Internal update of the solver runtime parameters.

virtual int	internal_update_params_sparse (unsigned int stage, int idx, double p, int n_update)=0
	A sparse implementation of `internal_update_params_sparse()`.

virtual int	internal_solve ()=0
	Solve the non-linear optimization`.

virtual int	internal_simulate (double dt, double x0, double u0, double p, double x_next, double *z_next)=0
	Simulate the next state given the current state, control inputs and runtime parameters.

virtual void	internal_print_stats () const =0
	Print the solver stats (e.g., number of SQP iters, number of QP iters, etc.) to the console.

Protected Attributes
Dimensions	_dims
	Fixed dimensions of the imported Acados OCP.

IndexMap	_x_index_map
	Unordered map used to set or retrieve the values of diff. state variables by name.

IndexMap	_z_index_map
	Unordered map used to set or retrieve the values of algebraic state variables by name.

IndexMap	_p_index_map
	Unordered map used to set the values of runtime parameters by name.

IndexMap	_u_index_map
	Unordered map used to set or retrieve the values of control variables by name.

Detailed Description

Abstract C++ wrapper of generated Acados solver C-code.

Constructor & Destructor Documentation

◆ AcadosSolver()

acados::AcadosSolver::AcadosSolver ( )

Constructor of the AcadosSolver object.

Does nothing for now.

Warning: Any class inheriting from AcadosSolver will have to correctly initialize the AcadosSolver::_dims object.

◆ ~AcadosSolver()

virtual acados::AcadosSolver::~AcadosSolver ( )

virtual

Destructor of the AcadosSolver object.

Warning: The free_memory() function must called by the destructor of any class inheriting from AcadosSolver. Otherwise, the allocated memory won't be fully released.

Member Function Documentation

◆ init()

int acados::AcadosSolver::init	(	unsigned int	N,
		double	Ts
	)

Initialize the solver and set the (constant) sampling intervals.

Todo:: Add alternative init() function for non-constant sampling intervals.

Parameters

N	Number of shooting nodes (strictly positive).
Ts	Sampling time in seconds.

Returns: int (zero if all OK).

◆ reset()

int acados::AcadosSolver::reset ( )

Resets the Acados solver, including the internal QP solver and the RTI phase cache.

Returns: int (zero if all OK).

◆ solve()

int acados::AcadosSolver::solve ( )

Solve the non-linear optimization problem given the provided initial state, constraints, etc.

The method returns the internal Acados status.

Returns: 0 : All OK.; 1 : Invalid input (NaN detected).; 2 : Maximum number of iteration reached.; 3 : Minimum step length reached, could possibly be a NaN solution too (if using HPIPM); 4 : QP solver failed.; other : Not OK.

◆ solve_rti()

int acados::AcadosSolver::solve_rti ( RtiStage rti_phase )

Solve the non-linear (RTI) optimization problem given the provided initial state, constraints, etc.

The method returns the internal Acados status.

Parameters

rti_phase The RTI phase to be solved, i.e., either RtiStage::PREPARATION or RtiStage::FEEDBACK.

Returns: 0 : All OK.; 1 : Invalid input (NaN detected).; 2 : Maximum number of iteration reached.; 3 : Minimum step length reached, could possibly be a NaN solution too (if using HPIPM); 4 : QP solver failed.; other : Not OK.

◆ simulate() [1/3]

int acados::AcadosSolver::simulate	(	double	dt,
		ValueVector &	x0,
		ValueVector &	u0,
		ValueVector &	p,
		ValueVector &	x_next,
		ValueVector &	z
	)

Simulate the next state given the current state, control inputs and runtime parameters.

The method returns the internal Acados status.

Returns: 0 : All OK.; other : Not OK. See Acados documentation for details.

◆ simulate() [2/3]

int acados::AcadosSolver::simulate	(	double	dt,
		ValueMap const &	x0_map,
		ValueMap const &	u0_map,
		ValueMap const &	p_map,
		ValueMap &	x_next_map,
		ValueMap &	z_map
	)

Simulate the next state given the current state, control inputs and runtime parameters.

See the other simulate() method for details.

◆ simulate() [3/3]

int acados::AcadosSolver::simulate	(	ValueMap const &	x0_map,
		ValueMap const &	u0_map,
		ValueMap const &	p_map,
		ValueMap &	x_next_map,
		ValueMap &	z_map
	)

Simulate the next state given the current state, control inputs and runtime parameters.

See the other simulate() method for details.

◆ set_initial_state_values() [1/2]

int acados::AcadosSolver::set_initial_state_values ( ValueVector & x_0 )

Set the initial state values (i.e., add constraints on initial state)

Parameters

x_0	State values at initial stage.

Returns: int (zero if all OK)

◆ set_initial_state_values() [2/2]

int acados::AcadosSolver::set_initial_state_values ( ValueMap const & x_0_map )

Set the initial state values (i.e., add constraints on initial state) from a map of key/ValueVector pairs.

Parameters

x_0_map ValueMap of the initial state values.

Returns: int (zero if all OK)

◆ set_state_bounds() [1/2]

int acados::AcadosSolver::set_state_bounds	(	unsigned int	stage,
		IndexVector &	idxbx,
		ValueVector &	lbx,
		ValueVector &	ubx
	)

Set (differential) state bounds at a given stage.

Parameters

stage	Stage in [0;N].
idxbx	Indexes of the bounded state variables.
lbx	Vector of lower bounds.
ubx	Vector of uower bounds.

Returns: int (zero if all OK)

◆ set_state_bounds() [2/2]

int acados::AcadosSolver::set_state_bounds	(	IndexVector &	idxbx,
		ValueVector &	lbx,
		ValueVector &	ubx
	)

Set (differential) state bounds for all stages.

The bounds are therefore equal at each stage.

Parameters

idxbx	Indexes of the bounded state variables.
lbx	Vector of lower bounds.
ubx	Vector of uower bounds.

Returns: int (zero if all OK)

◆ set_control_bounds() [1/2]

int acados::AcadosSolver::set_control_bounds	(	unsigned int	stage,
		IndexVector &	idxbu,
		ValueVector &	lbu,
		ValueVector &	ubu
	)

Set control bounds at a given stage.

Parameters

stage	Stage in [0;N].
idxbu	Indexes of the bounded control variables.
lbu	Vector of lower bounds.
ubu	Vector of uower bounds.

Returns: int (zero if all OK)

◆ set_control_bounds() [2/2]

int acados::AcadosSolver::set_control_bounds	(	IndexVector &	idxbu,
		ValueVector &	lbu,
		ValueVector &	ubu
	)

Set control bounds for all stages.

The bounds are therefore equal at each stage.

Parameters

idxbu	Indexes of the bounded control variables.
lbu	Vector of lower bounds.
ubu	Vector of uower bounds.

Returns: int (zero if all OK)

◆ set_runtime_parameters() [1/4]

int acados::AcadosSolver::set_runtime_parameters	(	unsigned int	stage,
		ValueVector &	p_i
	)

Set the runtime parameters for one stage from ordered value vector.

Parameters

stage	Stage in [0;N].
p_i	Vector containing the (ordered) runtime parameters.

Returns: int Status (zero if all OK).

◆ set_runtime_parameters() [2/4]

int acados::AcadosSolver::set_runtime_parameters	(	unsigned int	stage,
		ValueMap const &	p_i_map
	)

Set the runtime parameters for one stage from a key/values map.

Note: If the parameters are the same for all stages, prefer the set_runtime_parameters(ValueMap const &) alternative.

Parameters

stage	Stage in [0;N].
p_i_map	Key to ValueVector map containing the runtime parameters.

Returns: int Status (zero if all OK).

◆ set_runtime_parameters() [3/4]

int acados::AcadosSolver::set_runtime_parameters ( ValueVector & p_i )

Set the runtime parameters for all stages at once from ordered value vector.

The runtime parameters for ALL stages are equal.

Parameters

p_i	Vector containing the (ordered) runtime parameters.

Returns: int Status (zero if all OK).

◆ set_runtime_parameters() [4/4]

int acados::AcadosSolver::set_runtime_parameters ( ValueMap const & p_i_map )

Set the runtime parameters for all stages at once from a key/values map.

The runtime parameters for ALL stages are equal. Once the ordered vector of parameters is extracted, it is the set_runtime_parameters(unsigned int, ValueVector &) function that is used.

Parameters

p_i_map Key to ValueVector map containing the runtime parameters.

Returns: int Status (zero if all OK).

◆ initialize_state_values() [1/4]

int acados::AcadosSolver::initialize_state_values	(	unsigned int	stage,
		ValueVector &	x_i
	)

Initialize the (differential) state values for a stage from an ordered value vector.

Warning: This function (and other "initialization" functions similar to this one) only initialize the solver state/control values. These function should only be used before a AcadoSolver::solve() call! Also, initializing the state values for stage zero is not equivalent to set the initial state (see AcadoSolver::set_initial_state_values()).

Parameters

stage	Stage in [0;N].
x_i	Vector containing the (ordered) state values.

Returns: int Status (zero if all OK).

◆ initialize_state_values() [2/4]

int acados::AcadosSolver::initialize_state_values	(	unsigned int	stage,
		ValueMap const &	x_i_map
	)

Initialize the (differential) state values for a stage from a key/values map.

Note: If the parameters are the same for all stages, prefer the set_runtime_parameters(ValueMap const &) alternative.

Parameters

stage	Stage in [0;N].
x_i_map	Key to ValueVector map containing the initial state values.

Returns: int Status (zero if all OK).

◆ initialize_state_values() [3/4]

int acados::AcadosSolver::initialize_state_values ( ValueVector & x_i )

Initialize the (differential) state values for ALL stages at once from an ordered value vector.

Parameters

x_i	Vector containing the (ordered) state values.

Returns: int Status (zero if all OK).

◆ initialize_state_values() [4/4]

int acados::AcadosSolver::initialize_state_values ( ValueMap const & x_i_map )

Initialize the (differential) state values for ALL stages at once from a key/values map.

Parameters

x_i_map Key to ValueVector map containing the initial state values.

Returns: int Status (zero if all OK).

◆ initialize_control_values() [1/4]

int acados::AcadosSolver::initialize_control_values	(	unsigned int	stage,
		ValueVector &	u_i
	)

Initialize the control variable values for one stage from an ordered value vector.

Parameters

stage	Stage in [0;N].
u_i	Vector containing the (ordered) control variables values.

Returns: int Status (zero if all OK).

◆ initialize_control_values() [2/4]

int acados::AcadosSolver::initialize_control_values	(	unsigned int	stage,
		ValueMap const &	u_i_map
	)

Initialize the control variable values for one stage from a key/values map.

Parameters

stage	Stage in [0;N].
u_i_map	Key to ValueVector map containing the control variables values.

Returns: int Status (zero if all OK).

◆ initialize_control_values() [3/4]

int acados::AcadosSolver::initialize_control_values ( ValueVector & u_i )

Initialize the control variable values for ALL stages at once from an ordered value vector.

Parameters

u_i	Vector containing the (ordered) control variables values.

Returns: int Status (zero if all OK).

◆ initialize_control_values() [4/4]

int acados::AcadosSolver::initialize_control_values ( ValueMap const & u_i_map )

Initialize the control variable values for ALL stages at once from a key/values map.

Parameters

u_i_map Key to ValueVector map containing the control variables values.

Returns: int Status (zero if all OK).

◆ get_state_values()

ValueVector acados::AcadosSolver::get_state_values ( unsigned int stage )

Retrieve the differential state variables at a given stage.

Parameters

stage Stage in [0;N].

Returns: ValueVector Vector of (ordered) state values.

◆ get_state_values_as_map()

ValueMap acados::AcadosSolver::get_state_values_as_map ( unsigned int stage )

Retrieve the differential state variables at a given stage and package them as a ValueMap.

Parameters

stage Stage in [0;N].

Returns: ValueMap Key/ValueVector map containing of differential state values.

◆ get_algebraic_state_values()

ValueVector acados::AcadosSolver::get_algebraic_state_values ( unsigned int stage )

Retrieve the algebraic state variables at a given stage.

Parameters

stage Stage in [0;N].

Returns: ValueVector Vector of (ordered) state values.

◆ get_algebraic_state_values_as_map()

ValueMap acados::AcadosSolver::get_algebraic_state_values_as_map ( unsigned int stage )

Retrieve the algebraic state variables at a given stage and package them as a ValueMap.

Parameters

stage Stage in [0;N].

Returns: ValueMap Key/ValueVector map containing the algebraic state values.

◆ get_control_values()

ValueVector acados::AcadosSolver::get_control_values ( unsigned int stage )

Retrieve the control variables at a given stage.

Parameters

stage Stage in [0;N].

Returns: ValueVector Vector of (ordered) control variable values.

◆ get_control_values_as_map()

ValueMap acados::AcadosSolver::get_control_values_as_map ( unsigned int stage )

Retrieve the control variables at a given stage and package them as a ValueMap.

Parameters

stage Stage in [0;N].

Returns: ValueMap Key/ValueVector map containing the control variable values.

◆ get_parameter_values()

ValueVector acados::AcadosSolver::get_parameter_values ( unsigned int stage )

Retrieve the parameters at a given stage.

Parameters

stage Stage in [0;N].

Returns: ValueMap Key/ValueVector map containing the parameters values.

◆ get_parameter_values_as_map()

ValueMap acados::AcadosSolver::get_parameter_values_as_map ( unsigned int stage )

Retrieve the parameters at a given stage and package them as a ValueMap.

Parameters

stage Stage in [0;N].

Returns: ValueMap Key/ValueVector map containing the parameters values.

◆ x_index_map()

const IndexMap & acados::AcadosSolver::x_index_map ( ) const

Returns a constant reference to the Key/IndexVector map of differential state variables internally stored in AcadosSolver::_x_index_map.

Returns: const IndexMap&

◆ z_index_map()

const IndexMap & acados::AcadosSolver::z_index_map ( ) const

Returns a constant reference to the Key/IndexVector map of algebraic state variables internally stored in AcadosSolver::_z_index_map.

Returns: const IndexMap&

◆ p_index_map()

const IndexMap & acados::AcadosSolver::p_index_map ( ) const

Returns a constant reference to the Key/IndexVector map of runtime parameters internally stored in AcadosSolver::_p_index_map.

Returns: const IndexMap&

◆ u_index_map()

const IndexMap & acados::AcadosSolver::u_index_map ( ) const

Returns a constant reference to the Key/IndexVector map of control variables internally stored in AcadosSolver::_u_index_map.

Returns: const IndexMap&

◆ is_map_size_consistent() [1/2]

static bool acados::AcadosSolver::is_map_size_consistent	(	IndexMap const &	index_map,
		unsigned int	expected_total_indexes
	)

static

Check if the total number of indexes contained in the provided index map is equal to an expected number.

Parameters

index_map	The Key/IndexVector map to be tested for size consistency.
expected_total_indexes	The expected number of indexes.

Returns: true if the size is consistent; false otherwise

◆ is_map_size_consistent() [2/2]

static bool acados::AcadosSolver::is_map_size_consistent	(	ValueMap const &	values_map,
		unsigned int	expected_total_indexes
	)

static

Check if the total number of values contained in the provided value map is equal to an expected number.

Parameters

values_map	The Key/ValueVector map to be tested for size consistency.
expected_total_indexes	The expected number of values.

Returns: true if the size is consistent; false otherwise

◆ is_values_map_complete()

static bool acados::AcadosSolver::is_values_map_complete	(	IndexMap const &	index_map,
		ValueMap const &	values_map
	)

static

Check if for each key-index pair contained in the index map there is a value provided by the value map.

It is considered complete even if there are too many values (e.g., unused keys in values_map).

Parameters

index_map	The Key/IndexVector map to be tested for size consistency.
values_map	The Key/ValueVector map to be tested for size consistency.

Returns: true if the value map is complete.; false otherwise.

◆ fill_vector_from_map()

static void acados::AcadosSolver::fill_vector_from_map	(	IndexMap const &	index_map,
		ValueMap const &	values_map,
		unsigned int	vector_size,
		ValueVector &	values
	)

static

Fill (and resize if necessary) a vector with the values contained in the value map.

The index map is used to arrange the values within the vector.

Exceptions

std::invalid_argument

Todo:: Overloaded function without the vector_size argument. The vector_size argument could (should?) be optional since it can be retrieved from the number of indexes contained in index_map.

Parameters

[in]	index_map	Mapping between keys and indexes.
[in]	values_map	Mapping between keys and values to write in the value vector.
[in]	vector_size	The expected size of the resulting value vector.
[out]	values	The output container where the values will be written.

◆ value_vector_from_map()

static ValueVector acados::AcadosSolver::value_vector_from_map	(	IndexMap const &	index_map,
		ValueMap const &	values_map,
		unsigned int	vector_size
	)

static

Returns a vector containing the values provided through the value map.

Parameters

[in]	index_map	Mapping between keys and indexes.
[in]	values_map	Mapping between keys and values to write in the value vector.
[out]	vector_size	The expected size of the resulting value vector.

Returns: ValueVector The resulting value vector.

◆ create_map_from_values()

static ValueMap acados::AcadosSolver::create_map_from_values	(	IndexMap const &	index_map,
		ValueVector const &	values
	)

static

Returns a vector ValueMap object containing the values of the value vector.

Warning: The values must have been encoded as described by the index_map argument.

Parameters

[in]	index_map	Mapping between keys and indexes.
[in]	values	Vector of (ordered) values.

◆ fill_map_from_values()

static void acados::AcadosSolver::fill_map_from_values	(	IndexMap const &	index_map,
		ValueVector const &	values,
		ValueMap &	value_map
	)

static

Fill a ValueMap object with the values of the value vector.

Warning: You have to ensure that the index_map and values are consistent, i.e., that the number of indexes in index_map is equal to the size of values.

Parameters

[in]	index_map	Mapping between keys and indexes.
[in]	values	Vector of (ordered) values.
[out]	value_map	The ValueMap object to be filled with the values.

◆ dims()

const Dimensions & acados::AcadosSolver::dims ( ) const

Returns the (fixed) dimensions of the OCP problem stored in AcadosSolver::_dims.

Returns: const Dimensions& The dimensions

◆ nx()

unsigned int acados::AcadosSolver::nx ( ) const

Returns the dimensions of the differential state vector.

This is a convenience wrapper of the call dims().nx (see dims()).

Returns: unsigned int nx

◆ nz()

unsigned int acados::AcadosSolver::nz ( ) const

Returns the dimensions of the algebraic state vector.

This is a convenience wrapper of the call dims().nz (see dims()).

Returns: unsigned int nz

◆ np()

unsigned int acados::AcadosSolver::np ( ) const

Returns the number of runtime parameters.

This is a convenience wrapper of the call dims().np (see dims()).

Returns: unsigned int np

◆ nu()

unsigned int acados::AcadosSolver::nu ( ) const

Returns the number of control variables.

This is a convenience wrapper of the call dims().nu (see dims()).

Returns: unsigned int nu

◆ N()

unsigned int acados::AcadosSolver::N ( ) const

Returns the number of shooting nodes.

Returns: unsigned int N

◆ Ts()

double acados::AcadosSolver::Ts ( ) const

Returns the (constant) sampling interval Ts (in seconds).

If the sampling intervals are of variable duration (see sampling_intervals()), the function returns -1.

Returns: unsigned int Ts

◆ sampling_intervals()

std::vector< double > acados::AcadosSolver::sampling_intervals ( ) const

Returns the vector of sampling interval (in seconds).

Returns: std::vector<double> the vector of sampling intervals

◆ create_index_maps()

virtual int acados::AcadosSolver::create_index_maps ( )

protectedpure virtual

Reinitialize the maps used to set or retrieve value using string variable names.

◆ free_memory()

int acados::AcadosSolver::free_memory ( )

protected

Free all the allocated memory.

Triggers calls to internal_free(), internal_free_capsule(), etc.

Warning: Should be called by the destructor of any class inheriting from AcadosSolver.

Returns: int (zero if all OK).

◆ internal_create_capsule()

virtual int acados::AcadosSolver::internal_create_capsule ( )

protectedpure virtual

Create the Acados solver's capsule used by the Acados C-interface.

Wrapper of the <acados_model_name>_acados_create_capsule() C function. Note that the storage of the capsule itself is handled by the inherited class.

Returns: int (zero if all OK).

◆ internal_create_with_discretization()

virtual int acados::AcadosSolver::internal_create_with_discretization	(	int	n_time_steps,
		double *	new_time_steps
	)

protectedpure virtual

Initialize Acados solver and (optionally) redefine the shooting nodes.

This is a wrapper of the <acados_model_name>_acados_create_with_discretization(...) C function. For constant sampling intervals, provide new_time_steps = nullptr as argument.

Parameters

n_time_steps	number of shooting nodes.
new_time_steps	C-array of sampling intervals (in seconds).

Returns: int (zero if all OK).

◆ internal_reset()

virtual int acados::AcadosSolver::internal_reset ( int reset_qp_solver_mem = 1 )

protectedpure virtual

Reset the solver memory and (opt.) the internal QP solver.

This is a wrapper of the <acados_model_name>_acados_reset(...) C function.

Parameters

reset_qp_solver_mem Reset if =1.

Returns: int (zero if all OK).

◆ internal_free()

virtual int acados::AcadosSolver::internal_free ( )

protectedpure virtual

Free the memory allocated for the solver.

This is a wrapper of the <acados_model_name>_acados_free(...) C function. Call internal_free() before internal_free_capsule().

Returns: int (zero if all OK).

◆ internal_free_capsule()

virtual int acados::AcadosSolver::internal_free_capsule ( )

protectedpure virtual

Free the memory allocated for the capsule.

This is a wrapper of the <acados_model_name>_acados_free_capsule(...) C function.

Returns: int (zero if all OK).

◆ internal_update_qp_solver_cond_N()

virtual int acados::AcadosSolver::internal_update_qp_solver_cond_N ( int qp_solver_cond_N )

protectedpure virtual

Update the QP solver qp_solver_cond_N setting.

This is a wrapper of the <acados_model_name>_acados_update_qp_solver_cond_N(...) C function.

Parameters

qp_solver_cond_N New horizon after partial condensing.

Returns: int (zero if all OK).

◆ internal_update_params()

virtual int acados::AcadosSolver::internal_update_params	(	unsigned int	stage,
		double *	value,
		int	np
	)

protectedpure virtual

Internal update of the solver runtime parameters.

This is a wrapper of the <acados_model_name>_acados_update_params(...) C function.

Parameters

stage	Stage in [0;N].
value	Parameter values (C-array).
np	Number of parameters.

Returns: int (zero if all OK).

◆ internal_update_params_sparse()

virtual int acados::AcadosSolver::internal_update_params_sparse	(	unsigned int	stage,
		int *	idx,
		double *	p,
		int	n_update
	)

protectedpure virtual

A sparse implementation of internal_update_params_sparse().

This is a wrapper of the <acados_model_name>_acados_update_params_sparse(...) C function.

Parameters

stage	Stage in [0;N].
idx	Parameter indexes (C-array).
p	Parameter values (C-array).
n_update	Number of parameters to update.

Returns: int (zero if all OK).

◆ internal_solve()

virtual int acados::AcadosSolver::internal_solve ( )

protectedpure virtual

Solve the non-linear optimization`.

This is a wrapper of the <acados_model_name>_acados_solve(...) C function.

Returns: int (zero if all OK). See Acados documentation for a list of return flags.

◆ internal_simulate()

virtual int acados::AcadosSolver::internal_simulate	(	double	dt,
		double *	x0,
		double *	u0,
		double *	p,
		double *	x_next,
		double *	z_next
	)

protectedpure virtual

Simulate the next state given the current state, control inputs and runtime parameters.

This is a wrapper of the <acados_model_name>_acados_solve(...) C function. The method returns the internal Acados status.

Returns: int (zero if all OK). See Acados documentation for a list of return flags.

◆ internal_print_stats()

virtual void acados::AcadosSolver::internal_print_stats ( ) const

protectedpure virtual

Print the solver stats (e.g., number of SQP iters, number of QP iters, etc.) to the console.

This is a wrapper of the <acados_model_name>_acados_print_stats(...) C function.

◆ get_nlp_in()

virtual ocp_nlp_in * acados::AcadosSolver::get_nlp_in ( ) const

pure virtual

Legacy C-function to retrieve the C-interface ocp_nlp_in property of the solver.

This is a wrapper of the auto-generated <acados_model_name>_acados_get_nlp_in(...) C function.

Returns: ocp_nlp_in* Pointer towards the ocp_nlp_in object.

◆ get_nlp_out()

virtual ocp_nlp_out * acados::AcadosSolver::get_nlp_out ( ) const

pure virtual

Legacy C-function to retrieve the C-interface ocp_nlp_out property of the solver.

This is a wrapper of the auto-generated <acados_model_name>_acados_get_nlp_out(...) C function.

Returns: ocp_nlp_out* Pointer towards the ocp_nlp_out object.

◆ get_sens_out()

virtual ocp_nlp_out * acados::AcadosSolver::get_sens_out ( ) const

pure virtual

Legacy C-function to retrieve the C-interface ocp_nlp_out property of the solver.

This is a wrapper of the auto-generated <acados_model_name>_acados_get_sens_out(...) C function. The function is used to retrieve the sensibility matrices.

Returns: ocp_nlp_out* Pointer towards the ocp_nlp_out object.

◆ get_nlp_solver()

virtual ocp_nlp_solver * acados::AcadosSolver::get_nlp_solver ( ) const

pure virtual

Legacy C-function to retrieve the C-interface ocp_nlp_solver property of the solver.

This is a wrapper of the auto-generated <acados_model_name>_acados_get_nlp_solver(...) C function.

Returns: ocp_nlp_solver* Pointer towards the ocp_nlp_solver object.

◆ get_nlp_config()

virtual ocp_nlp_config * acados::AcadosSolver::get_nlp_config ( ) const

pure virtual

Legacy C-function to retrieve the C-interface ocp_nlp_config property of the solver.

This is a wrapper of the auto-generated <acados_model_name>_acados_get_nlp_config(...) C function.

Returns: ocp_nlp_config* Pointer towards the ocp_nlp_config object.

◆ get_nlp_opts()

virtual void * acados::AcadosSolver::get_nlp_opts ( ) const

pure virtual

Legacy C-function to retrieve the C-interface get_nlp_opts property of the solver.

This is a wrapper of the auto-generated <acados_model_name>_acados_get_nlp_opts(...) C function.

Returns: void* Pointer towards the nlp_opts object.

◆ get_nlp_dims()

virtual ocp_nlp_dims * acados::AcadosSolver::get_nlp_dims ( ) const

pure virtual

Legacy C-function to retrieve the C-interface ocp_nlp_dims property of the solver that contains the dimensions of the problem.

This is a wrapper of the auto-generated <acados_model_name>_acados_get_nlp_dims(...) C function.

Returns: ocp_nlp_dims* Pointer towards the ocp_nlp_dims object.

◆ get_nlp_plan()

virtual ocp_nlp_plan_t * acados::AcadosSolver::get_nlp_plan ( ) const

pure virtual

Legacy C-function to retrieve the C-interface ocp_nlp_plan property of the solver.

This is a wrapper of the auto-generated <acados_model_name>_acados_get_nlp_plan(...) C function.

Returns: ocp_nlp_plan_t* Pointer towards the ocp_nlp_plan object.

◆ get_nlp_np()

virtual unsigned int acados::AcadosSolver::get_nlp_np ( ) const

pure virtual

Legacy C-function to retrieve the number of runtime parameters from the Acados capsule.

This function returns the field capsule->nlp_np where capsule is the internal pointer to the acados solver data.

Returns: unsigned int np

Member Data Documentation

◆ _dims

Dimensions acados::AcadosSolver::_dims

protected

Fixed dimensions of the imported Acados OCP.

◆ _x_index_map

IndexMap acados::AcadosSolver::_x_index_map

protected

Unordered map used to set or retrieve the values of diff. state variables by name.

◆ _z_index_map

IndexMap acados::AcadosSolver::_z_index_map

protected

Unordered map used to set or retrieve the values of algebraic state variables by name.

◆ _p_index_map

IndexMap acados::AcadosSolver::_p_index_map

protected

Unordered map used to set the values of runtime parameters by name.

◆ _u_index_map

IndexMap acados::AcadosSolver::_u_index_map

protected

Unordered map used to set or retrieve the values of control variables by name.

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

/home/runner/work/acados_solver_ros2/acados_solver_ros2/acados_solver_base/include/acados_solver_base/acados_solver.hpp

Classes

Public Member Functions

Static Public Member Functions

Protected Member Functions

Protected Attributes

Detailed Description

Constructor & Destructor Documentation

◆ AcadosSolver()

◆ ~AcadosSolver()

Member Function Documentation

◆ init()

◆ reset()

◆ solve()

◆ solve_rti()

◆ simulate() [1/3]

◆ simulate() [2/3]

◆ simulate() [3/3]

◆ set_initial_state_values() [1/2]

◆ set_initial_state_values() [2/2]

◆ set_state_bounds() [1/2]

◆ set_state_bounds() [2/2]

◆ set_control_bounds() [1/2]

◆ set_control_bounds() [2/2]

◆ set_runtime_parameters() [1/4]

◆ set_runtime_parameters() [2/4]

◆ set_runtime_parameters() [3/4]

◆ set_runtime_parameters() [4/4]

◆ initialize_state_values() [1/4]

◆ initialize_state_values() [2/4]

◆ initialize_state_values() [3/4]

◆ initialize_state_values() [4/4]

◆ initialize_control_values() [1/4]

◆ initialize_control_values() [2/4]

◆ initialize_control_values() [3/4]

◆ initialize_control_values() [4/4]

◆ get_state_values()

◆ get_state_values_as_map()

◆ get_algebraic_state_values()

◆ get_algebraic_state_values_as_map()

◆ get_control_values()

◆ get_control_values_as_map()

◆ get_parameter_values()

◆ get_parameter_values_as_map()

◆ x_index_map()

◆ z_index_map()

◆ p_index_map()

◆ u_index_map()

◆ is_map_size_consistent() [1/2]

◆ is_map_size_consistent() [2/2]

◆ is_values_map_complete()

◆ fill_vector_from_map()

◆ value_vector_from_map()

◆ create_map_from_values()

◆ fill_map_from_values()

◆ dims()

◆ nx()

◆ nz()

◆ np()

◆ nu()

◆ N()

◆ Ts()

◆ sampling_intervals()

◆ create_index_maps()

◆ free_memory()

◆ internal_create_capsule()

◆ internal_create_with_discretization()

◆ internal_reset()

◆ internal_free()

◆ internal_free_capsule()

◆ internal_update_qp_solver_cond_N()

◆ internal_update_params()

◆ internal_update_params_sparse()

◆ internal_solve()

◆ internal_simulate()

◆ internal_print_stats()

◆ get_nlp_in()

◆ get_nlp_out()

◆ get_sens_out()

◆ get_nlp_solver()

◆ get_nlp_config()