Use IMU measurements to set submodel base state in RDE

[isorrentino]

This PR implements what is described in #70 and #95. In addition, the PR introduces a code restructure to have the possibility to define more than one measurement dynamics for a given sensor. The tests have been modified accordingly.

[UPDATE] I added also the possibility of having the friction torques as a measurement for the estimator.

[isorrentino]

@GiulioRomualdi after the rebase the CI fails.

[isorrentino]

@traversaro I requested also your review to be sure about the implemented formulas. If you can check the code here https://github.com/ami-iit/bipedal-locomotion-framework/blob/b8163a9e1b7f1aad447cce7912501db3d3b3c22a/src/Estimators/src/UkfModel.cpp#L114 that is where I take the imu measurements to set the submodel base states and compute the forward dynamics as discussed in https://github.com/ami-iit/element_sensorless-torque-control/issues/70, it would be great :)

[traversaro]

@traversaro I requested also your review to be sure about the implemented formulas. If you can check the code here

https://github.com/ami-iit/bipedal-locomotion-framework/blob/b8163a9e1b7f1aad447cce7912501db3d3b3c22a/src/Estimators/src/UkfModel.cpp#L114 that is where I take the imu measurements to set the submodel base states and compute the forward dynamics as discussed in ami-iit/element_sensorless-torque-control#70, it would be great :)

Th computation seems correct. However, formula in https://github.com/ami-iit/bipedal-locomotion-framework/blob/b8163a9e1b7f1aad447cce7912501db3d3b3c22a/src/Estimators/src/UkfModel.cpp#L154 are not obvious, so if you could add a reference (for example with a comment) for that equation it would be great.

[GiulioRomualdi]

Hi @isorrentino can you rebase the PR on top of master?

[isorrentino]

Hi @isorrentino can you rebase the PR on top of master?

Done.