MockControlClosedLoop¶

class lsst.ts.m2com.MockControlClosedLoop(is_mirror: bool = False)¶

Bases: object

Mock closed-loop control.

Parameters¶

is_mirrorbool, optional: Is mirror or not. If not, the surrogate is applied. (the default is False)

Attributes¶

is_runningbool: The closed-loop control is running or not.
temperaturedict: Temperature of mirror in degree C.
axial_forcesdict: Forces of the axial actuators in Newton.
tangent_forcesdict: Forces of the tangent actuators in Newton.
hardpointslist: List of the hardpoints. There are 6 actuators. The first three are the axial actuators and the latter three are the tangent links.
disp_hardpoint_homelist: Displacement of the hardpoints at the home position. The unit is meter.
control_loopMockControlLoop: Mock control loop with the force control algorithm.

Methods Summary

`apply_forces`(force_axial, force_tangent)	Apply the actuator forces.
`calc_cmd_delay_filter_params`([is_mirror, ...])	Calculate the command delay filter parameters (or more clearly, the numerator of the transfer function).
`calc_cmd_prefilter_params`([numerator, ...])	Calculate the command pre-filter parameters in biquadratic filters.
`calc_force_control_filter_params`([...])	Calculate the force control filter parameters in biquadratic filters.
`calc_hp_comp_matrix`(location_axial_actuator, ...)	Calculate the hardpoint compensation matrix.
`calc_kinetic_decoupling_matrix`(...)	Calculate the kinetic decoupling matrix.
`calc_look_up_forces`([lut_angle, ...])	Calculate look-up table (LUT) forces using current system state (position and temperature).
`calc_temp_inv_matrix`()	Calculate the temperature inversion matrix.
`calculate_rigid_body_xy`(radius, ...)	Calculate the rigid body (x, y) position.
`get_active_actuators`(hardpoints_axial, ...)	Get the active actuators.
`get_actuator_location_axial`()	Get the location of axial actuators.
`get_actuator_location_tangent`()	Get the location of tangent actuators.
`get_demanded_force`([applied_force_axial, ...])	Get the demanded force in Newton.
`get_force_balance`()	Get the data of force balance system.
`get_net_forces_total`()	Get the total net forces in Newton.
`get_net_moments_total`()	Get the total net moments in Newton * meter.
`get_radius`()	Get the radius of the tangential actuators.
`handle_forces`(is_in_position[, plant, ...])	Handle forces and update the hardpoint correction and measured forces.
`hardpoint_to_rigid_body`(...)	Calculate the rigid body position based on the hardpoint displacements.
`is_actuator_force_out_limit`([...])	The actuator force is out of limit or not.
`is_cell_temperature_high`()	Cell temperature is high or not.
`load_file_cell_geometry`(filepath)	Load the file of cell geometry.
`load_file_lut`(filepath)	Load the look-up table (LUT) files.
`load_file_stiffness`(filepath)	Load the file of stiffness matrix.
`reset_force_offsets`()	Reset the force offsets.
`rigid_body_to_actuator_displacement`(...)	Calculate the actuator displacements based on the rigid body position.
`set_hardpoint_compensation`()	Set the hardpoint compensation matrix.
`set_kinetic_decoupling_matrix`()	Set the kinetic decoupling matrix.
`set_measured_forces`(force_axial, force_tangent)	Set the measured actuator forces.
`simulate_temperature_and_update`([...])	Simulate the temperature change and update the internal value.
`transfer_function_to_biquadratic_filter`(...)	Transfer function to the biquadratic filters.
`update_hardpoints`(hardpoints)	Update the hardpoints and related internal parameters.

Methods Documentation

apply_forces(force_axial: list[float] | ndarray[Any, dtype[float64]], force_tangent: list[float] | ndarray[Any, dtype[float64]]) → None¶

Apply the actuator forces.

Parameters¶

force_axiallist or numpy.ndarray: Force of the axial actuators in Newton.
force_tangentlist or numpy.ndarray: Force of the tangent actuators in Newton.

Raises¶

RuntimeError: When the maximum force limits reached.

static calc_cmd_delay_filter_params(is_mirror: bool = True, control_frequency: float = 20.0, bypass_delay: bool = False, num_degree: int = 5) → list[float]¶

Calculate the command delay filter parameters (or more clearly, the numerator of the transfer function).

Notes¶

Translate the calculation from the Create_Filter_Params.m in ts_mtm2_matlab_tools.

For a transfer function, H(z), if the numerator is [1, 3, 3] and the denominator is [1, 2, 1], it will be:

H(z) = (z^2 + 3 * z + 3) / (z^2 + 2 * z + 1)

Since the denominator of transfer function is the demanded force change here, the linear transfer ([1.0]) is applied. Therefore, we only care about the delay of command as the output.

Parameters¶

is_mirrorbool, optional: Is mirror or not. If not, the surrogate is applied. (the default is True)
control_frequencyfloat, optional: Frequency of the control loop in Hz. (the default is 20.0)
bypass_delaybool, optional: Bypass the command delay or not. (the default is False)
num_degreeint, optional: Number of the degree in the transfer function. (the default is 5)

Returns¶

numerator_transfer_functionlist: Numerator of the transfer function from the high degree to low degree.

static calc_cmd_prefilter_params(numerator: list[float] | None = None, denominator: list[float] | None = None) → tuple[float, numpy.ndarray[Any, numpy.dtype[numpy.float64]]]¶

Calculate the command pre-filter parameters in biquadratic filters.

Notes¶

Translate the calculation from the Create_Filter_Params.m in ts_mtm2_matlab_tools.

Currently, pre-filter is a pass-through filter.

Parameters¶

numeratorlist or None, optional: Numerator polynomial coefficients of the transfer function. If None, the linear transfer is assumed (aka. [1.0]). (the default is None)
denominatorlist or None, optional: Denominator polynomial coefficients of the transfer function. If None, the linear transfer is assumed (aka. [1.0]). (the default is None)

Returns¶

float: Gain.
list: Coefficients of the command pre-filter parameters.

static calc_force_control_filter_params(is_mirror: bool = True, numerator: list[float] | None = None, denominator: list[float] | None = None) → tuple[float, numpy.ndarray[Any, numpy.dtype[numpy.float64]]]¶

Calculate the force control filter parameters in biquadratic filters.

Notes¶

Translate the calculation from the Create_Filter_Params.m in ts_mtm2_matlab_tools.

Parameters¶

is_mirrorbool, optional: Is mirror or not. If not, the surrogate is applied. (the default is True)
numeratorlist or None, optional: Numerator polynomial coefficients of the transfer function. If None, the linear transfer is assumed (aka. [1.0]). (the default is None)
denominatorlist or None, optional: Denominator polynomial coefficients of the transfer function. If None, the linear transfer is assumed (aka. [1.0]). (the default is None)

Returns¶

float: Gain.
coefficientslist: Coefficients of the force control filter parameters.

static calc_hp_comp_matrix(location_axial_actuator: list[list], hardpoints_axial: list[int], hardpoints_tangent: list[int]) → tuple[numpy.ndarray[Any, numpy.dtype[numpy.float64]], numpy.ndarray[Any, numpy.dtype[numpy.float64]]]¶

Calculate the hardpoint compensation matrix.

Notes¶

Translate the calculation from the CalcHPFCInfMat.m in ts_mtm2_matlab_tools.

The axial hardpoint compensation matrix (M) fulfills:

M * (x_hp, y_hp, 1) = (x_nhp, y_nhp, 1)

x_hp: x-position of hardpoint y_hp: y-position of hardpoint x_nhp: x-position of non-hardpoint y_nhp: y-position of non-hardpoint

The idea is to make the x-moment amd y-moment keep the same when distributes the force of axial hardpoints to other axial actuators. It is the same idea for tangential actuators with z-moment.

Assume the hardpoints are A1, A3, and A5. If they were active, the z-moment would be: 3 * fm * mirror_radius, where fm = (f1 + f3 + f5) / 3.

For the active tagent links: A2, A4, and A5 to maintain this z-moment while keeping the hardpoints to be passive, we have:

f2 = -(f5 - fm) + fm = 2 * fm - f5 = 2 / 3 * (f1 + f3 + f5) - f5 f4 = -(f1 - fm) + fm = 2 * fm - f1 = 2 / 3 * (f1 + f3 + f5) - f1 f6 = -(f3 - fm) + fm = 2 * fm - f3 = 2 / 3 * (f1 + f3 + f5) - f3

Then, we have:

f2 = 2 / 3 * f1 + 2 / 3 * f3 - 1 / 3 * f5 f4 = -1 / 3 * f1 + 2 / 3 * f3 + 2 / 3 * f5 f6 = 2 / 3 * f1 - 1 / 3 * f3 + 2 / 3 * f5

In the matrix form, it will be:

(f2, f4, f6).T = M * (f1, f3, f5).T

If the hardpoints were A2, A4, A6, we would have:

(f1, f3, f5).T = M ^ (-1) * (f2, f4, f6).T

Parameters¶

location_axial_actuatorlist: Location of the axial actuators: (x, y). This should be a 72 x 2 matrix.
hardpoints_axiallist: Three axial hardpoints. The order is from low to high, e.g. [5, 15, 25].
hardpoints_tangentlist: Three tangential hardpoints. This can only be [72, 74, 76] or [73, 75, 77]. The order is from low to high.

Returns¶

hd_comp_axialnumpy.ndarray: Axial hardpoint compensation matrix.
hd_comp_tangentnumpy.ndarray: Tangential hardpoint compensation matrix.

static calc_kinetic_decoupling_matrix(location_axial_actuator: list[list], hardpoints_axial: list[int], hardpoints_tangent: list[int], stiffness_matrix: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]¶

Calculate the kinetic decoupling matrix.

Notes¶

Translate the calculation from the CalcHPFCInfMat.m in ts_mtm2_matlab_tools.

The unit of the element is the motor step (could be fraction, row) per Newton (column). This matrix is used to multiply with the desired actuator’s force changes to get the related amounts of motor’s step changes.

delta_force = M * delta_step => M^-1 = kdc to have: delta_step = kdc * delta_force

Parameters¶

location_axial_actuatorlist: Location of the axial actuators: (x, y). This should be a 72 x 2 matrix.
hardpoints_axiallist: Three axial hardpoints. The order is from low to high, e.g. [5, 15, 25].
hardpoints_tangentlist: Three tangential hardpoints. This can only be [72, 74, 76] or [73, 75, 77]. The order is from low to high.
stiffness_matrixnumpy.ndarray: Stiffness matrix of M2 mirror or surrogate.

Returns¶

numpy.ndarray: Kinetic decoupling matrix.

calc_look_up_forces(lut_angle: float | None = None, enable_lut_temperature: bool | None = None) → None¶

Calculate look-up table (LUT) forces using current system state (position and temperature).

Parameters¶

lut_anglefloat or None, optional: Angle used to calculate the LUT forces of gravity component. If None, bypass the calculation. (the default is None)
enable_lut_temperaturebool or None, optional: Enable the temperature LUT calculation or not. If None, bypass the calculation. (the default is None)

static calc_temp_inv_matrix() → ndarray[Any, dtype[float64]]¶

Calculate the temperature inversion matrix.

Based on the “LSST_M2_temperature_sensors_20171102.pdf”. The sensor map is at “doc/figure/temperature_sensor_map.jpg”.

Radial gradient: T(r) = Tr * r / R X gradient: T(x) = Tx * x / R = Tx * r * cos(theta) / R Y gradient: T(y) = Ty * y / R = Ty * r * sin(theta) / R Uniform bias: T(u) = Tu

The matrix is: [r/R, r/R * cos(theta), r/R * sin(theta), 1], and we have: matrix * [Tr, Tx, Ty, Tu].T = [T1, T2, …, T12].T

Returns¶

numpy.ndarray: Temperature inversion matrix.

static calculate_rigid_body_xy(radius: float, tangent_hardpoint_displacement: list[float], tangent_hardpoint_location: list[float]) → tuple[float, float, numpy.ndarray[Any, numpy.dtype[numpy.float64]]]¶

Calculate the rigid body (x, y) position.

Notes¶

This is part of the calculation in HardPointToRigidBody.vi in ts_mtm2.

Parameters¶

radiusfloat: Radius of the cell in meter.
tangent_hardpoint_displacementlist: Displacement of the 3 tangent hardpoints in meter.
tangent_hardpoint_locationlist: Location of the 3 tangent hardpoints in radian.

Returns¶

mean_xfloat: X position in meter.
mean_yfloat: Y position in meter.
delta_xy_tangent_hardpointnumpy.ndarray: Delta (x, y) position compared with the mean (x, y) in meter.

static get_active_actuators(hardpoints_axial: list[int], hardpoints_tangent: list[int]) → tuple[list[int], list[int]]¶

Get the active actuators.

Parameters¶

hardpoints_axiallist: Three axial hardpoints. The order is from low to high, e.g. [5, 15, 25].
hardpoints_tangentlist: Three tangential hardpoints. This can only be [72, 74, 76] or [73, 75, 77]. The order is from low to high.

Returns¶

active_actuators_axiallist: Axial active actuators.
active_actuators_tangentlist: Tangential active actuators.

Raises¶

ValueError: If the tangential hardpoints are wrong.

get_actuator_location_axial() → list¶

Get the location of axial actuators.

Returns¶

list: Location (x, y) of the axial actuators in meter.

get_actuator_location_tangent() → list¶

Get the location of tangent actuators.

Returns¶

list: Location of the tangential actuators in degree.

get_demanded_force(applied_force_axial: ndarray[Any, dtype[float64]] | None = None, applied_force_tangent: ndarray[Any, dtype[float64]] | None = None) → ndarray[Any, dtype[float64]]¶

Get the demanded force in Newton.

Parameters¶

applied_force_axialnumpy.ndarray or None, optional: Axial forces to apply in Newton. If None use current setup. (the default is None.)
applied_force_tangentnumpy.ndarray or None, optional: Tangent forces to apply in Newton. If None use current setup. ( the default is None.)

Returns¶

numpy.ndarray: Array with the combined total force per actuator in Newton.

get_force_balance() → dict¶

Get the data of force balance system. This contains the net forces (in Newton) and net moments (in Newton * meter).

Returns¶

force_balancedict: Data of the force balance system.

get_net_forces_total() → dict¶

Get the total net forces in Newton.

Returns¶

dict: Total net forces in Newton.

get_net_moments_total() → dict¶

Get the total net moments in Newton * meter.

Returns¶

dict: Total net moments in Newton * meter.

get_radius() → float¶

Get the radius of the tangential actuators.

Returns¶

float: Radius of the tangential actuators in meter.

handle_forces(is_in_position: bool, plant: MockPlant | None = None, steps_hardpoints: ndarray[Any, dtype[int64]] | None = None) → bool¶

Handle forces and update the hardpoint correction and measured forces.

Parameters¶

is_in_positionbool: Mirror is in position or not.
plantMockPlant or None: Plant model. If not None, the movement of plant model will be applied.
steps_hardpointsnumpy.ndarray [int] or None: Hardpoint steps of the rigid body movement. (the default is None)

Returns¶

in_positionbool: M2 assembly is in position or not.

static hardpoint_to_rigid_body(location_axial_actuator: list[list], location_tangent_link: list[float], radius: float, hardpoints: list[int], disp_hardpoint_current: list[float], disp_hardpoint_home: list[float]) → tuple[float, float, float, float, float, float]¶

Calculate the rigid body position based on the hardpoint displacements.

Notes¶

Translate the calculation from the HardPointToRigidBody.vi in ts_mtm2.

Parameters¶

location_axial_actuatorlist [list]: Location of the axial actuators: (x, y) in meter. This should be a (NUM_ACTUATOR - NUM_TANGENT_LINK) x 2 matrix.
location_tangent_linklist: Location of the tangent links in degree. This should be a 1 x NUM_TANGENT_LINK array.
radiusfloat: Radius of the cell in meter.
hardpointslist: Six hardpoints. The order is from low to high.
disp_hardpoint_currentlist: Six current hardpoint displacements. The unit is meter.
disp_hardpoint_homelist: Six hardpoint displacements at the home position. The unit is meter.

Returns¶

xfloat: X position in meter.
yfloat: Y position in meter.
zfloat: Z position in meter.
rxfloat: X rotator in radian.
ryfloat: Y rotator in radian.
rzfloat: Z rotator in radian.

is_actuator_force_out_limit(applied_force_axial: ndarray[Any, dtype[float64]] | None = None, applied_force_tangent: ndarray[Any, dtype[float64]] | None = None, use_measured_force: bool = False) → tuple[bool, list, list]¶

The actuator force is out of limit or not.

Notes¶

Remove the ‘use_measured_force’ after we implement the forward modeling in the control algorithm. At that time, the measured force should be consistent with the demanded force (with the consideration of hardpoint correction).

Parameters¶

applied_force_axialnumpy.ndarray or None, optional: Axial forces to apply in Newton. If None use current setup. (the default is None.)
applied_force_tangentnumpy.ndarray or None, optional: Tangent forces to apply in Newton. If None use current setup. ( the default is None.)
use_measured_forcebool, optional: Use the measured force to compare with the limit. (the default is False)

Returns¶

bool: True if the actuator force is out of limit. Otherwise, False.
list: Triggered retracted limit switches.
list: Triggered extended limit switches.

is_cell_temperature_high() → bool¶

Cell temperature is high or not.

Returns¶

bool: True if the cell temperature is high. Otherwise, False.

load_file_cell_geometry(filepath: Path) → None¶

Load the file of cell geometry.

Parameters¶

filepathpathlib.PosixPath: File path of cell geometry.

load_file_lut(filepath: Path) → None¶

Load the look-up table (LUT) files.

Parameters¶

filepathpathlib.PosixPath: File path of LUT directory.

load_file_stiffness(filepath: Path) → None¶

Load the file of stiffness matrix.

Parameters¶

filepathpathlib.PosixPath: File path of stiffness matrix.

reset_force_offsets() → None¶

Reset the force offsets.

This will put the applied force to be zero.

static rigid_body_to_actuator_displacement(location_axial_actuator: list[list], location_tangent_link: list[float], radius: float, dx: float, dy: float, dz: float, drx: float, dry: float, drz: float) → ndarray[Any, dtype[float64]]¶

Calculate the actuator displacements based on the rigid body position.

Notes¶

Translate the calculation from the RigidBodyToActuatorDisplacement.vi in ts_mtm2. For the “radius”, the original developer had the following comment:

An average of 5 of the 6 tangent location radius from the center of the M2 cell which is to be used by the rigid body to displacement calculation. Technically, it is not the B-ring radius and this should be changed in the future.

Parameters¶

location_axial_actuatorlist [list]: Location of the axial actuators: (x, y) in meter. This should be a (NUM_ACTUATOR - NUM_TANGENT_LINK) x 2 matrix.
location_tangent_linklist: Location of the tangent links in degree. This should be a 1 x NUM_TANGENT_LINK array.
radiusfloat: Radius of the cell in meter.
dxfloat: Delta x position in meter.
dyfloat: Delta y position in meter.
dzfloat: Delta z position in meter.
drxfloat: Delta x rotator in radian.
dryfloat: Delta y rotator in radian.
drzfloat: Delta z rotator in radian.

Returns¶

numpy.ndarray: All actuator displacements in meter.

set_hardpoint_compensation() → None¶: Set the hardpoint compensation matrix.

set_kinetic_decoupling_matrix() → None¶: Set the kinetic decoupling matrix.

set_measured_forces(force_axial: ndarray[Any, dtype[float64]], force_tangent: ndarray[Any, dtype[float64]]) → None¶

Set the measured actuator forces.

Parameters¶

force_axialnumpy.ndarray: Axial force in Newton.
force_tangentnumpy.ndarray: Tangential force in Newton.

Raises¶

ValueError: When the input dimensions of forces are wrong.

simulate_temperature_and_update(temperature_rms: float = 0.05, max_temperature: float = 28.0, min_temperature: float = 0.0) → None¶

Simulate the temperature change and update the internal value.

Parameters¶

temperature_rmsfloat, optional: Temperature rms variation in degree C. (the default is 0.05)
max_temperaturefloat, optional: Maximum temperature in degree C. (the default is 28.0)
min_temperaturefloat, optional: Minimum temperature in degree C. (the default is 0.0)

static transfer_function_to_biquadratic_filter(numerator: list[float], denominator: list[float], num_stage: int = 8) → tuple[float, numpy.ndarray[Any, numpy.dtype[numpy.float64]]]¶

Transfer function to the biquadratic filters.

Notes¶

Translate the calculation from the bqd.m and sos2pqd.m in ts_mtm2_matlab_tools.

The reference is: https://en.wikipedia.org/wiki/Digital_biquad_filter

Compact biquadratic format:

1 + b11 z^-1 + b21 z^-2 1 + b12 z^-1 + b22 z^-2

H(z) = gain * (———————–) (———————–) …

1 + a11 z^-1 + a21 z^-2 1 + a12 z^-1 + a22 z^-2

1 + b1N z^-1 + b2N z^-2

(———————–): 1 + a1N z^-1 + a2N z^-2

Format of the coefficent output is: [a11 a21 b11 b21 a12 a22 b12 b22 … a1N a2N b1N b2N]

Parameters¶

numeratorlist: Numerator polynomial coefficients of the transfer function.
denominatorlist: Denominator polynomial coefficients of the transfer function.
num_stageint, optional: Number of the stage of biquadratic filters. (the default is 8)

Returns¶

gainfloat: Gain.
coefficientslist: Coefficients of the biquadratic filters.

update_hardpoints(hardpoints: list[int]) → None¶

Update the hardpoints and related internal parameters.

Parameters¶

hardpointslist: List of the 0-based hardpoints. There are 6 actuators. The first three are the axial actuators and the latter three are the tangent links.

Navigation

MockControlClosedLoop¶

Parameters¶

Attributes¶

Parameters¶

Raises¶

Notes¶

Parameters¶

Returns¶

Notes¶

Parameters¶

Returns¶

Notes¶

Parameters¶

Returns¶

Notes¶

Parameters¶

Returns¶

Notes¶

Parameters¶

Returns¶

Parameters¶

Returns¶

Notes¶

Parameters¶

Returns¶

Parameters¶

Returns¶

Raises¶

Returns¶

Returns¶

Parameters¶

Returns¶

Returns¶

Returns¶

Returns¶

Returns¶

Parameters¶

Returns¶

Notes¶

Parameters¶

Returns¶

Notes¶

Parameters¶

Returns¶

Returns¶

Parameters¶

Parameters¶

Parameters¶

Notes¶

Parameters¶

Returns¶

Parameters¶

Raises¶

Parameters¶

Notes¶

Parameters¶

Returns¶

Parameters¶