Skip to content
Snippets Groups Projects
Commit 6bbe59c8 authored by Joan Solà Ortega's avatar Joan Solà Ortega
Browse files

Improve doc and remove old commented code

parent cd1acf05
No related branches found
No related tags found
2 merge requests!31devel->main,!30Complete UAV identification setup
...@@ -177,12 +177,9 @@ void ProcessorForceTorqueInertialDynamics::data2tangent(const Eigen::VectorXd& _ ...@@ -177,12 +177,9 @@ void ProcessorForceTorqueInertialDynamics::data2tangent(const Eigen::VectorXd& _
const Matrix3d& fd_cross = skew(fd); const Matrix3d& fd_cross = skew(fd);
const Matrix3d& c_cross = skew(c); const Matrix3d& c_cross = skew(c);
// tangent(a,w) = data(a,w) - bias(a,w) _tangent.segment<6>(0) = awd - _bias; // awt = awd - awb
// tangent(f) = data(f) _tangent.segment<3>(6) = fd; // ft = fd
// tangent(t) = data(t) - model(c) x data(f) _tangent.segment<3>(9) = td - c.cross(fd); // tt = td - c x fd
_tangent.segment<6>(0) = awd - _bias;
_tangent.segment<3>(6) = fd;
_tangent.segment<3>(9) = td - c.cross(fd); // c x fd
// J_tangent_data // J_tangent_data
_J_tangent_data.setIdentity(12, 12); _J_tangent_data.setIdentity(12, 12);
...@@ -295,15 +292,52 @@ void ProcessorForceTorqueInertialDynamics::computeCurrentDelta(const Eigen::Vect ...@@ -295,15 +292,52 @@ void ProcessorForceTorqueInertialDynamics::computeCurrentDelta(const Eigen::Vect
* J_delta_tangent * J_delta_tangent
* J_delta_model * J_delta_model
* *
* these are obtained directly by the function g() = tangent2delta(). * these are obtained directly by the function g() = tangent2delta(). This function does:
* *
* Note 1: It is unclear to me (JS, 4-aug-2022) if this function tangent2delta() is completely accurate * angacc = I.inv * ( tt - wt x ( I * wt ) )
* with regard to the two different reference frames: BASE and COM. It is possible that * --> J_aa_i, J_aa_tt, J_aa_wt
* we have to revise the math. * acc_dyn = ft / _mass - angacc x com - wt x ( wt x com )
* * --> J_ad_ft, J_ad_m, J_ad_aa, J_ad_c, J_ad_wt
* Note 2: It is even possible that the function tangent2delta() is OK, but then that the function * ==> J_ad_wt, J_ad_ft, J_ad_tt, J_ad_c, J_ad_i, J_ad_m
* betweenStates() does not account for the difference in reference frames. So we also need to revise *
* the math inside betweenStates() with regard to the two reference frames BASE and COM. * delta_p_imu = 0.5 * at * dt^2 --> J_dpi_at
* delta_v_imu = at * _dt --> J_dvi_at
* delta_p_dyn = 0.5 * acc_dyn * dt^2 --> J_dpd_ad
* delta_v_dyn = acc_dyn * _dt --> J_dvd_ad
* delta_L = tt * _dt --> J_dL_tt
* delta_q = exp_q(wt * _dt) --> J_dq_wt
*
* Assembling Jacobians gives:
*
* at wt ft tt
* J_delta_tangent = [ J_dpi_at 0 0 0
* J_dvi_at 0 0 0
* 0 J_dpd_wt J_dpd_ft J_dpd_tt
* 0 J_dvd_wt J_dvd_ft J_dvd_tt
* 0 0 0 J_dL_tt
* 0 J_dq_wt 0 0 ]
* with
* J_dpd_wt = J_dpd_ad * ( J_ad_wt + J_ad_aa * J_aa_wt )
* J_dpd_ft = J_dpd_ad * J_ad_ft
* J_dpd_tt = J_dpd_ad * J_ad_aa * J_aa_tt
* J_dvd_wt = J_dvd_ad * ( J_ad*wt + J_ad_aa * J_aa_wt )
* J_dvd_ft = J_dvd_ad * J_ad_ft
* J_dvd_tt = J_dvd_ad * J_ad_aa * J_aa_tt
*
* c i m
* J_delta_model = [ 0 0 0
* 0 0 0
* J_dpd_c J_dpd_i J_dpd_m
* J_dvd_c J_dvd_i J_dvd_m
* 0 0 0
* 0 0 0 ]
* with
* J_dpd_c = J_dpd_ad * J_ad_c
* J_dpd_i = J_dpd_ad * J_ad_aa * J_aa_i
* J_dpd_m = J_dpd_ad * J_ad_m
* J_dvd_c = J_dvd_ad * J_ad_c
* J_dvd_i = J_dvd_ad * J_ad_aa * J_aa_i
* J_dvd_m = J_dvd_ad * J_ad_m
* *
* 3) On a third stage, we need to apply the chain rule for the functions f() and g(), * 3) On a third stage, we need to apply the chain rule for the functions f() and g(),
* that is delta = g( f( data, bias, model), model) -- yes, it's a mess!! * that is delta = g( f( data, bias, model), model) -- yes, it's a mess!!
...@@ -317,10 +351,12 @@ void ProcessorForceTorqueInertialDynamics::computeCurrentDelta(const Eigen::Vect ...@@ -317,10 +351,12 @@ void ProcessorForceTorqueInertialDynamics::computeCurrentDelta(const Eigen::Vect
* *
* J_delta_calib = [ J_delta_bias | J_delta_model ] * J_delta_calib = [ J_delta_bias | J_delta_model ]
* *
*
* 4) On a fourth stage, we compute the covariance of the delta: * 4) On a fourth stage, we compute the covariance of the delta:
* *
* Q_delta = J_delta_data * Q_data * J_delta_data.transpose() * Q_delta = J_delta_data * Q_data * J_delta_data.transpose()
* *
*
* 5) The function returns the following quantities, which relate to the variables used above: * 5) The function returns the following quantities, which relate to the variables used above:
* *
* - _delta = delta * - _delta = delta
...@@ -346,26 +382,6 @@ void ProcessorForceTorqueInertialDynamics::computeCurrentDelta(const Eigen::Vect ...@@ -346,26 +382,6 @@ void ProcessorForceTorqueInertialDynamics::computeCurrentDelta(const Eigen::Vect
* J_(axb)_b = a_x * J_(axb)_b = a_x
*/ */
// Matrix<double, 12, 1> tangent = _data;
// tangent.head<6>() -= _calib.head<6>(); // J_tangent_data = Identity(12,12)
// Matrix<double, 12, 6> J_tangent_I; // J_tangent_I = [-Identity(6,6) ; Zero(6,6)]
// J_tangent_I.topRows<6>() = -Matrix6d::Identity();
// J_tangent_I.bottomRows<6>() = Matrix6d::Zero();
// // go from tangent to delta. We need to dynamic model for this
// const Matrix<double, 7, 1>& model = sensor_fti_->getModel();
// Matrix<double, 18, 12> J_delta_tangent;
// Matrix<double, 18, 7> J_delta_model;
// fti::tangent2delta(tangent, model, _dt, _delta, J_delta_tangent, J_delta_model);
// // Compute cov and compose jacobians
// Matrix<double, 18, 6> J_delta_I = J_delta_tangent * J_tangent_I;
// _jacobian_calib << J_delta_I, J_delta_model; // J_delta_calib = _jacobian_calib = [J_delta_I ; J_delta_model]
// const auto& J_delta_data = J_delta_tangent; // * J_tangent_data, which is the Identity;
// _delta_cov = J_delta_data * _data_cov * J_delta_data.transpose();
////////////////////// NOU CODI QUE HAURE DE REVISAR I CANVIAR PEL QUE HI HA ADALT //////////////////////////
/* /*
* 1. tangent = f(data, bias, model) --> J_tangent_data, J_tangent_bias, J_tangent_model * 1. tangent = f(data, bias, model) --> J_tangent_data, J_tangent_bias, J_tangent_model
* *
...@@ -465,13 +481,13 @@ bool ProcessorForceTorqueInertialDynamics::voteForKeyFrame() const ...@@ -465,13 +481,13 @@ bool ProcessorForceTorqueInertialDynamics::voteForKeyFrame() const
// time span // time span
if (getBuffer().back().ts_ - getBuffer().front().ts_ > params_force_torque_inertial_dynamics_->max_time_span) if (getBuffer().back().ts_ - getBuffer().front().ts_ > params_force_torque_inertial_dynamics_->max_time_span)
{ {
WOLF_DEBUG("PM: vote: time span"); WOLF_DEBUG("PM ", getName(), " vote: time span");
return true; return true;
} }
// buffer length // buffer length
if (getBuffer().size() > params_force_torque_inertial_dynamics_->max_buff_length) if (getBuffer().size() > params_force_torque_inertial_dynamics_->max_buff_length)
{ {
WOLF_DEBUG("PM: vote: buffer length"); WOLF_DEBUG("PM ", getName(), " vote: buffer length");
return true; return true;
} }
// dist_traveled // dist_traveled
...@@ -482,14 +498,14 @@ bool ProcessorForceTorqueInertialDynamics::voteForKeyFrame() const ...@@ -482,14 +498,14 @@ bool ProcessorForceTorqueInertialDynamics::voteForKeyFrame() const
double dist = (X1.at('P') - X0.at('P')).norm(); double dist = (X1.at('P') - X0.at('P')).norm();
if (dist > params_force_torque_inertial_dynamics_->dist_traveled) if (dist > params_force_torque_inertial_dynamics_->dist_traveled)
{ {
WOLF_DEBUG("PM: vote: distance traveled"); WOLF_DEBUG("PM ", getName(), " vote: distance traveled");
return true; return true;
} }
// angle turned // angle turned
double angle = 2.0 * acos(getMotion().delta_integr_(15)); double angle = 2.0 * acos(getMotion().delta_integr_(15));
if (angle > params_force_torque_inertial_dynamics_->angle_turned) if (angle > params_force_torque_inertial_dynamics_->angle_turned)
{ {
WOLF_DEBUG("PM: vote: angle turned"); WOLF_DEBUG("PM ", getName(), " vote: angle turned");
return true; return true;
} }
......
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment