#include <feature-posture.hh>
Public Types | |
typedef dynamicgraph::SignalPtr < dg::Vector, int > | signalIn_t |
typedef dynamicgraph::SignalTimeDependent < dg::Vector, int > | signalOut_t |
Public Member Functions | |
FeaturePosture (const std::string &name) | |
virtual | ~FeaturePosture () |
virtual unsigned int & | getDimension (unsigned int &res, int) |
Verbose method. | |
void | setPosture (const dg::Vector &posture) |
void | selectDof (unsigned dofId, bool control) |
Public Attributes | |
DECLARE_NO_REFERENCE | |
Protected Member Functions | |
virtual dg::Vector & | computeError (dg::Vector &res, int) |
Compute the error between the desired feature and the current value of the feature measured or deduced from the robot state. | |
virtual dg::Matrix & | computeJacobian (dg::Matrix &res, int) |
Compute the Jacobian of the error according the robot state. | |
virtual dg::Vector & | computeActivation (dg::Vector &res, int) |
Protected Attributes | |
signalIn_t | state_ |
signalIn_t | posture_ |
signalOut_t | error_ |
dg::Matrix | jacobian_ |
Friends | |
class | SelectDof |
Feature that observes the posture of the robot, ie whose Jacobian is the identity, or slices of the identity. This feature can be exactly obtained with a generic posture, given the identity matrix as the input Jacobian, the identity matrix. It is even prefereable, as the reference value is then given by a signal, which can be reevalutated at each iteration, for example to track a reference trajectory in the configuration space. See for example the toFlag python function in the sot-dyninv module to nicely selec the posture DOF.
typedef dynamicgraph::SignalTimeDependent<dg::Vector, int> dynamicgraph::sot::FeaturePosture::signalOut_t |
dynamicgraph::sot::FeaturePosture::FeaturePosture | ( | const std::string & | name | ) | [explicit] |
virtual dynamicgraph::sot::FeaturePosture::~FeaturePosture | ( | ) | [virtual] |
virtual dg::Vector& dynamicgraph::sot::FeaturePosture::computeActivation | ( | dg::Vector & | res, |
int | |||
) | [protected, virtual] |
virtual dg::Vector& dynamicgraph::sot::FeaturePosture::computeError | ( | dg::Vector & | res, |
int | time | ||
) | [protected, virtual] |
Compute the error between the desired feature and the current value of the feature measured or deduced from the robot state.
Implements dynamicgraph::sot::FeatureAbstract.
virtual dg::Matrix& dynamicgraph::sot::FeaturePosture::computeJacobian | ( | dg::Matrix & | res, |
int | time | ||
) | [protected, virtual] |
Compute the Jacobian of the error according the robot state.
Implements dynamicgraph::sot::FeatureAbstract.
virtual unsigned int& dynamicgraph::sot::FeaturePosture::getDimension | ( | unsigned int & | res, |
int | time | ||
) | [virtual] |
Verbose method.
Implements dynamicgraph::sot::FeatureAbstract.
void dynamicgraph::sot::FeaturePosture::selectDof | ( | unsigned | dofId, |
bool | control | ||
) |
void dynamicgraph::sot::FeaturePosture::setPosture | ( | const dg::Vector & | posture | ) |
friend class SelectDof [friend] |
signalOut_t dynamicgraph::sot::FeaturePosture::error_ [protected] |
signalIn_t dynamicgraph::sot::FeaturePosture::state_ [protected] |