Redesign the dcmotor controller: setpoint inputs, torque-space gains.
The dcmotor input block is any subset of the canonical list [pos, vel, ff, voltage], selected with input="pos vel ff voltage" and recorded as mjtCtrlInput bits in actuator_ctrlspec like pid. Tokens are required in canonical order: the attribute denotes a set, the block always packs canonically, and accepting permutations invites reading the string as a layout choice. The mode flag in gainprm[8] is retired (reserved, written 0). Controller gains are now in torque space, as for pid: the controller commands tau = kp*(q*-l) + kd*(v*-ldot) + ki*x_I + tau_ff over the present inputs (absent setpoints frozen at zero) and converts to drive voltage V = R/K * tau + K*ldot. The second term compensates back-EMF, as the current loop of a real torque-mode driver does (torque commands are current commands): commanded torque is delivered exactly until a limit binds, and the torque-speed envelope emerges from the Vmax clamp. The map uses the nameplate R: thermal resistance growth is not compensated, so a hot motor under-delivers by R/R(T). A stateless setpoint dcmotor now matches <pid> exactly, for any K and R; the old back-EMF droop remains available as the physical behavior of the raw voltage path. Voltage-space datasheet gains convert by K/R. Controller inputs require a positive motor constant (the map divides by K), and controller gains require a controller input. ff and voltage are distinct inputs, different in kind: ff is a torque feedforward added to the controller output, uniform with pid's ff (feedforward in the actuator's output space), while voltage is the raw terminal voltage of the physical device, injected downstream of the controller and its Vmax clamp, unclamped (ctrlrange bounds it if desired). input="voltage" is the default: the plain voltage-commanded motor, whose behavior is unchanged by this commit. The integrator always accumulates position error; the old velocity mode's integral term, ki*(int(u)dt - theta), which tracked the integral of the velocity command, is retired without replacement, keeping ki mode-independent -- commanded integrated velocity belongs to an integrator activation state, not to controller gains. slewmax rate-limits the first controller input -- position setpoint (rad/s), velocity setpoint (rad/s^2) or torque feedforward (N*m/s), each a real driver feature (reference ramping, ramped-velocity and ramped-torque input modes); the raw voltage input is never rate-limited and slewmax requires a controller input. input="none" selects the empty signature: the actuator owns no controls at all (nu = 0 is now legal with actuators present) and is purely passive -- LuGre friction, cogging and back-EMF braking as passive joint forces. This exists because auxiliary dynamic states (the LuGre bristle) attach to actuators, not joints. The terminal voltage is identically zero, i.e. a shorted motor (dynamic braking); motorconst=0 decouples the electrical branch. mjINPUT_NONE is a distinct enum value because ctrlspec = 0 means "unset, use the type default". History and delay require an input; the controller voltage override and input read in mj_fwdActuation are gated on a nonempty block. The analytic velocity derivative of the controller becomes dV/dw = -kd*R/K + K, whose second term cancels the back-EMF bias exactly: the net damping of an unclipped torque-mode motor is -kd, and of a voltage-mode or passive motor -K^2/R. Viewers label inputs via mj_actuatorInputName: pos, vel, ff, voltage. The dcmotor LaTeX design doc is updated accordingly: torque-space units, the tau->V map and its saturation-generated envelope, the input-block pipeline figure, and a Passive Operation section. PiperOrigin-RevId: 965795351 Change-Id: Ibc308ca21bd6bad014e77f950ee08feaad449b73
This commit is contained in:
committed by
Copybara-Service
parent
11fa4a5b45
commit
2f1843f4a7
+40
-19
@@ -5691,7 +5691,7 @@ specify them independently.
|
||||
|
||||
.. _actuator-general-dyntype:
|
||||
|
||||
:at:`dyntype`: :at-val:`[none, integrator, filter, filterexact, pid, muscle, user], "none"`
|
||||
:at:`dyntype`: :at-val:`[none, integrator, filter, filterexact, pid, dcmotor, muscle, user], "none"`
|
||||
Activation dynamics type for the actuator. The available dynamics types were already described in the :ref:`Actuation
|
||||
model <geActuation>` section. Repeating that description in somewhat different notation (corresponding to the mjModel
|
||||
and mjData fields involved) we have:
|
||||
@@ -5711,7 +5711,7 @@ specify them independently.
|
||||
|
||||
.. _actuator-general-gaintype:
|
||||
|
||||
:at:`gaintype`: :at-val:`[fixed, affine, muscle, pid, so3, user], "fixed"`
|
||||
:at:`gaintype`: :at-val:`[fixed, affine, muscle, dcmotor, pid, so3, user], "fixed"`
|
||||
The gain and bias together determine the output of the force generation mechanism, which is currently assumed to be
|
||||
affine. As already explained in :ref:`Actuation model <geActuation>`, the general formula is:
|
||||
scalar_force = gain_term \* (act or ctrl) + bias_term.
|
||||
@@ -5724,6 +5724,7 @@ specify them independently.
|
||||
fixed gain_term = gainprm[0]
|
||||
affine gain_term = gain_prm[0] + gain_prm[1]*length + gain_prm[2]*velocity
|
||||
muscle gain_term = mju_muscleGain(...)
|
||||
dcmotor DC motor gain (K or K/R), see :ref:`dcmotor<actuator-dcmotor>`
|
||||
pid PID controller with setpoint inputs, see :ref:`pid<actuator-pid>`
|
||||
so3 geodesic orientation servo, computed jointly over 3 force outputs, see :ref:`orientation<actuator-orientation>`
|
||||
user gain_term = mjcb_act_gain(...)
|
||||
@@ -5731,7 +5732,7 @@ specify them independently.
|
||||
|
||||
.. _actuator-general-biastype:
|
||||
|
||||
:at:`biastype`: :at-val:`[none, affine, muscle, so3, user], "none"`
|
||||
:at:`biastype`: :at-val:`[none, affine, muscle, dcmotor, so3, user], "none"`
|
||||
The keywords have the following meaning:
|
||||
|
||||
======= ================================================================
|
||||
@@ -5740,6 +5741,7 @@ specify them independently.
|
||||
none bias_term = 0
|
||||
affine bias_term = biasprm[0] + biasprm[1]*length + biasprm[2]*velocity
|
||||
muscle bias_term = mju_muscleBias(...)
|
||||
dcmotor DC motor bias: back-EMF, cogging, LuGre friction, see :ref:`dcmotor<actuator-dcmotor>`
|
||||
so3 damping term of the geodesic orientation servo, see :ref:`orientation<actuator-orientation>`
|
||||
user bias_term = mjcb_act_bias(...)
|
||||
======= ================================================================
|
||||
@@ -5784,8 +5786,9 @@ specify them independently.
|
||||
:at:`input`: :at-val:`string, optional`
|
||||
Input signature of the actuator: which controls make up its control block, recorded in
|
||||
``mjModel.actuator_ctrlspec``. For gaintype "so3" it selects the orientation chart: "expmap" (3 controls, the
|
||||
default) or "quat" (4 controls); see :ref:`orientation/input<actuator-orientation-input>`. For gaintype "pid" it is
|
||||
a token list selecting the input subset; see :ref:`pid/input<actuator-pid-input>`.
|
||||
default) or "quat" (4 controls); see :ref:`orientation/input<actuator-orientation-input>`. For gaintypes "pid" and
|
||||
"dcmotor" it is a token list selecting the input subset; see :ref:`pid/input<actuator-pid-input>` and
|
||||
:ref:`dcmotor/input<actuator-dcmotor-input>`.
|
||||
|
||||
.. _actuator-general-actearly:
|
||||
|
||||
@@ -6806,8 +6809,11 @@ the stateless case.
|
||||
- :ref:`resistance<actuator-dcmotor-resistance>`, :ref:`motorconst<actuator-dcmotor-motorconst>` and
|
||||
:ref:`nominal<actuator-dcmotor-nominal>` are each optional, but some combination of them is required.
|
||||
See Section 2.1 of the `technical note <_static/dcmotor.pdf>`__.
|
||||
- The control :ref:`input<actuator-dcmotor-input>` semantic is either the voltage applied to the motor terminals (the
|
||||
default), or a position or velocity target for a :ref:`PID controller<actuator-dcmotor-controller>`.
|
||||
- The control block is selected by :ref:`input<actuator-dcmotor-input>`: any subset of ``[pos, vel, ff]``, where
|
||||
``pos`` and ``vel`` are setpoint inputs to the on-board :ref:`PID controller<actuator-dcmotor-controller>` and
|
||||
``ff`` is a torque feedforward added to its output; the ``voltage`` input is the raw terminal voltage. The
|
||||
default is the plain voltage-commanded motor. With
|
||||
``input="none"`` the actuator has no control inputs at all and acts as a purely passive device.
|
||||
- Optional features include electrical dynamics (:ref:`inductance<actuator-dcmotor-inductance>`),
|
||||
:ref:`cogging torque<actuator-dcmotor-cogging>`, :ref:`thermal resistance variation<actuator-dcmotor-thermal>`, and
|
||||
:ref:`LuGre<actuator-dcmotor-lugre>` friction.
|
||||
@@ -6946,22 +6952,37 @@ This element has the following custom attributes in addition to the common attri
|
||||
|
||||
.. _actuator-dcmotor-input:
|
||||
|
||||
:at:`input`: :at-val:`[voltage, position, velocity], "voltage"`
|
||||
Specifies the input signal semantics. In "voltage" mode, the control directly sets applied motor voltage. In
|
||||
"position" or "velocity" modes, the :ref:`PID controller<actuator-dcmotor-controller>` uses the control as a
|
||||
reference setpoint relative to the joint trajectory. (see `tech note <_static/dcmotor.pdf>`__, Section 2.5)
|
||||
:at:`input`: :at-val:`string, "voltage"`
|
||||
Input signature: a space-separated subset of the tokens "pos", "vel", "ff" and "voltage", required in this
|
||||
canonical order. The ``pos`` and ``vel`` inputs are setpoints for the on-board
|
||||
:ref:`controller<actuator-dcmotor-controller>`, and ``ff`` is a torque feedforward added to its output, as for
|
||||
:ref:`pid/input<actuator-pid-input>`. The ``voltage`` input is different in kind: it is the raw terminal voltage
|
||||
of the physical device, applied downstream of the controller and its :at-val:`Vmax` clamp.
|
||||
``input="voltage"`` (the default) is the plain voltage-commanded motor. Absent setpoint inputs are fixed at zero.
|
||||
The keyword "none" selects the empty signature: the actuator has no control inputs and is purely
|
||||
passive, useful for modeling :ref:`friction<actuator-dcmotor-lugre>` and :ref:`cogging<actuator-dcmotor-cogging>`
|
||||
as passive joint forces. The terminal voltage is zero, so back-EMF drives current through the (shorted) motor and
|
||||
brakes the joint; setting :ref:`motorconst<actuator-dcmotor-motorconst>` to zero disables the electrical branch.
|
||||
(see `tech note <_static/dcmotor.pdf>`__, Section 2.5)
|
||||
|
||||
.. _actuator-dcmotor-controller:
|
||||
|
||||
:at:`controller`: :at-val:`real(6), "0 0 0 0 0 0"`
|
||||
PID controller parameters, defined as :at:`controller` = ":at-val:`kp` :at-val:`ki` :at-val:`kd`
|
||||
:at-val:`slewmax` :at-val:`Imax` :at-val:`Vmax`". Depending on the :at:`input` mode, the controller stabilizes
|
||||
either position or velocity. If the :at:`input` mode is voltage, :at-val:`kp`, :at-val:`ki`, :at-val:`kd` are
|
||||
ignored. :at-val:`Vmax` sets the maximum drive voltage :math:`v_{\max}` (Volt); in position/velocity modes it clamps
|
||||
the controller output, in voltage mode it clamps the control signal (if :at:`ctrlrange` is also set, the tighter
|
||||
limit wins). A value of 0 (the default) disables the respective feature. When positive, :at-val:`slewmax` limits the
|
||||
setpoint rate-of-change, :at-val:`Imax` clamps the integrator state (anti-windup), and :at-val:`Vmax` clamps the
|
||||
drive voltage. (see `tech note <_static/dcmotor.pdf>`__, Section 2.5)
|
||||
:at-val:`slewmax` :at-val:`Imax` :at-val:`Vmax`". The gains are in torque space, as for
|
||||
:ref:`pid<actuator-pid>`: the controller commands the torque
|
||||
:math:`\tau = k_p (u_{pos} - l) + k_d (u_{vel} - \dot{l}) + k_i x_I + u_{f\!f}` over the inputs present in the
|
||||
:ref:`input<actuator-dcmotor-input>` signature, absent setpoints being fixed at zero, and drives the voltage
|
||||
:math:`v = (R/K)\,\tau + K \dot{l}`, the second term compensating back-EMF as in a current-controlled driver:
|
||||
commanded torque is delivered exactly until a limit is reached. Torque-space gains from datasheet voltage-space
|
||||
values are obtained by multiplying by :math:`K/R`. The integrator state
|
||||
:math:`x_I` accumulates position error and requires the ``pos`` input; controller gains require a controller
|
||||
input and a positive :ref:`motorconst<actuator-dcmotor-motorconst>`.
|
||||
A value of 0 (the default) disables the respective feature. When positive, :at-val:`slewmax` limits the
|
||||
rate-of-change of the first input (position setpoint in rad/s, or with signatures lacking ``pos``, velocity
|
||||
setpoint or torque feedforward), :at-val:`Imax` clamps the integrator state (anti-windup), and :at-val:`Vmax`
|
||||
clamps the drive voltage :math:`v_{\max}` (Volt), upstream of the raw ``voltage`` input.
|
||||
(see `tech note <_static/dcmotor.pdf>`__, Section 2.5)
|
||||
|
||||
.. _actuator-plugin:
|
||||
|
||||
@@ -6982,7 +7003,7 @@ Associate this actuator with an :ref:`engine plugin<exPlugin>`. Either :at:`plug
|
||||
|
||||
.. _actuator-plugin-dyntype:
|
||||
|
||||
:at:`dyntype`: :at-val:`[none, integrator, filter, filterexact, pid, muscle, user], "none"`
|
||||
:at:`dyntype`: :at-val:`[none, integrator, filter, filterexact, pid, dcmotor, muscle, user], "none"`
|
||||
Activation dynamics type for the actuator. The available dynamics types were already described in the :ref:`Actuation
|
||||
model <geActuation>` section. If :ref:`dyntype<actuator-general-dyntype>` is not "none", an activation variable will
|
||||
be added to the actuator. This variable will be added after any activation state computed by the plugin (see
|
||||
|
||||
Reference in New Issue
Block a user