autogenu-jupyter
An automatic code generator and the continuation/GMRES (C/GMRES) based numerical solvers for nonlinear MPC
|
Settings of solvers. More...
#include <solver_settings.hpp>
Public Member Functions | |
void | disp (std::ostream &os) const |
Public Attributes | |
size_t | max_iter = 100 |
Maximum number of iterations of the ZeroHorizonOCPSolver method. Has nothing to do with SingleShootingCGMRESSolver or MultipleShootingCGMRESSolver. Default value is 100. More... | |
Scalar | opterr_tol = 1.0e-04 |
Termination criterion of the ZeroHorizonOCPSolver method. Has nothing to do with SingleShootingCGMRESSolver or MultipleShootingCGMRESSolver. Must be non-negative. Default value is 1.0e-04. More... | |
Scalar | finite_difference_epsilon = 1.0e-08 |
Epsilon of the finite difference approximation. Must be positive. Default value is 1.0e-08. More... | |
Scalar | sampling_time = 0.001 |
The sampling time of MPC and used in SingleShootingCGMRESSolver and MultipleShootingCGMRESSolver. Has nothing to do with ZeroHorizonOCPSolver. Must be positive. Default is 0.001. More... | |
Scalar | zeta = 1000.0 |
The stabilization parameter of the continuation method. Typical value is the reciprocal of SolverSettings::sampling_time (sampling period of MPC). Used in SingleShootingCGMRESSolver and MultipleShootingCGMRESSolver. Has nothing to do with ZeroHorizonOCPSolver. Must be positive. Default is 1000.0. More... | |
Scalar | min_dummy = 1.0e-03 |
The minimum value of the dummy inputs. Mainly used in MultipleShootingCGMRESSolver. In SingleShootingCGMRESSolver and ZeroHorizonOCPSolver, this value is used only in SingleShootingCGMRESSolver::init_dummy_mu() and ZeroHorizonOCPSolver::init_dummy_mu(). Must be non-negaive. Default is 1.0e-03. More... | |
size_t | verbose_level = 0 |
Verbose level. 0: no printings. 1-2: print some things. Default is 0. More... | |
bool | profile_solver = true |
If true, a solver profile is taken. More... | |
Friends | |
std::ostream & | operator<< (std::ostream &os, const SolverSettings &settings) |
Settings of solvers.
|
inline |
|
friend |
Scalar cgmres::SolverSettings::finite_difference_epsilon = 1.0e-08 |
Epsilon of the finite difference approximation. Must be positive. Default value is 1.0e-08.
size_t cgmres::SolverSettings::max_iter = 100 |
Maximum number of iterations of the ZeroHorizonOCPSolver method. Has nothing to do with SingleShootingCGMRESSolver or MultipleShootingCGMRESSolver. Default value is 100.
Scalar cgmres::SolverSettings::min_dummy = 1.0e-03 |
The minimum value of the dummy inputs. Mainly used in MultipleShootingCGMRESSolver. In SingleShootingCGMRESSolver and ZeroHorizonOCPSolver, this value is used only in SingleShootingCGMRESSolver::init_dummy_mu() and ZeroHorizonOCPSolver::init_dummy_mu(). Must be non-negaive. Default is 1.0e-03.
Scalar cgmres::SolverSettings::opterr_tol = 1.0e-04 |
Termination criterion of the ZeroHorizonOCPSolver method. Has nothing to do with SingleShootingCGMRESSolver or MultipleShootingCGMRESSolver. Must be non-negative. Default value is 1.0e-04.
bool cgmres::SolverSettings::profile_solver = true |
If true, a solver profile is taken.
Scalar cgmres::SolverSettings::sampling_time = 0.001 |
The sampling time of MPC and used in SingleShootingCGMRESSolver and MultipleShootingCGMRESSolver. Has nothing to do with ZeroHorizonOCPSolver. Must be positive. Default is 0.001.
size_t cgmres::SolverSettings::verbose_level = 0 |
Verbose level. 0: no printings. 1-2: print some things. Default is 0.
Scalar cgmres::SolverSettings::zeta = 1000.0 |
The stabilization parameter of the continuation method. Typical value is the reciprocal of SolverSettings::sampling_time (sampling period of MPC). Used in SingleShootingCGMRESSolver and MultipleShootingCGMRESSolver. Has nothing to do with ZeroHorizonOCPSolver. Must be positive. Default is 1000.0.