Python data processing blocks

`qmt.Block`()	Base class for online data processing blocks.
`qmt.OriEstVQFBlock`(Ts)	VQF orientation estimation algorithm.
`qmt.OriEstMadgwickBlock`(Ts)	Madgwicks's orientation estimation algorithm.
`qmt.OriEstMahonyBlock`(Ts)	Mahony's orientation estimation algorithm.
`qmt.OriEstIMUBlock`(Ts)	OriEstIMU orientation estimation algorithm.
`qmt.LowpassFilterBlock`(Ts[, fc, tau])	Second-order Butterworth low-pass filter block.

class qmt.Block[source]

Base class for online data processing blocks.

command(command)[source]

This function is called to handle commands, i.e., events that change the state of the block in a certain way.

Commands are always lists and the first entry is the command name. One common example for a command is ['reset']. The default implementation ignores all commands.

Parameters:: command – Command as list. The first element is a string with the command name. Optional further elements depend on the command.
Returns:: True if the command is handled by the block, False if the command was ignored

setParams(params)[source]

This function is called to update parameters, i.e., variables that influence how the input data is processed.

The default implementation updates the params property if a corresponding key already exists and raises a RuntimeError on other parameters (unless ignoreUnknownParams is set to True). If this is not enough, this function can be overridden to implement custom logic if parameters change.

Parameters:: params – dictionary with the parameters to update
Returns:: None

step(*args, **kwargs)[source]

Performs one data processing step.

The default implementation just returns None.

Parameters:

args – application-specific inputs
kwargs – application-specific inputs

Returns:

application-specific outputs

class qmt.OriEstVQFBlock(Ts)[source]

VQF orientation estimation algorithm.

See https://github.com/dlaidig/vqf, https://doi.org/10.1016/j.inffus.2022.10.014, and https://arxiv.org/abs/2203.17024 for more information about this algorithm.

This algorithm is also available as a function: qmt.oriEstVQF().

Parameters:: Ts – sampling time in seconds

command(command)[source]

This function is called to handle commands, i.e., events that change the state of the block in a certain way.

Commands are always lists and the first entry is the command name. One common example for a command is ['reset']. The default implementation ignores all commands.

Parameters:: command – Command as list. The first element is a string with the command name. Optional further elements depend on the command.
Returns:: True if the command is handled by the block, False if the command was ignored

setParams(params)[source]

This function is called to update parameters, i.e., variables that influence how the input data is processed.

The default implementation updates the params property if a corresponding key already exists and raises a RuntimeError on other parameters (unless ignoreUnknownParams is set to True). If this is not enough, this function can be overridden to implement custom logic if parameters change.

Parameters:: params – dictionary with the parameters to update
Returns:: None

step(gyr, acc, mag=None, debug=False)[source]

Performs one data processing step.

The default implementation just returns None.

Parameters:

args – application-specific inputs
kwargs – application-specific inputs

Returns:

application-specific outputs

class qmt.OriEstMadgwickBlock(Ts)[source]

Madgwicks’s orientation estimation algorithm.

See https://doi.org/10.1109/ICORR.2011.5975346 for more information about this algorithm. Based on the C++ implementation by Sebastian Madgwick, available at https://x-io.co.uk/open-source-imu-and-ahrs-algorithms/.

This algorithm is also available as a function: qmt.oriEstMadgwick().

Parameters:: Ts – sampling time in seconds

command(command)[source]

Supported commands:

[‘reset’]: resets the orientation estimation state

setParams(params)[source]

Supported parameters:

beta: algorithm gain

step(gyr, acc, mag, debug=False)[source]

Parameters:

gyr – gyroscope measurement [rad/s]
acc – accelerometer measurement [m/s^2]
mag – magnetometer measurement or None [any unit]
debug – enables debug output

Returns:

quat: orientation quaternion
debug: dict with debug values (only if debug==True)

class qmt.OriEstMahonyBlock(Ts)[source]

Mahony’s orientation estimation algorithm.

See https://dx.doi.org/10.1109/TAC.2008.923738 for more information about this algorithm. Based on the C++ implementation by Sebastian Madgwick, available at https://x-io.co.uk/open-source-imu-and-ahrs-algorithms/.

This algorithm is also available as a function: qmt.oriEstMahony().

Parameters:: Ts – sampling time in seconds

command(command)[source]

Supported commands:

[‘reset’]: resets the orientation estimation state

setParams(params)[source]

Supported parameters:

Kp: proportional gain
Ki: integral gain (for gyroscope bias estimation)

step(gyr, acc, mag, debug=False)[source]

Parameters:

gyr – gyroscope measurement [rad/s]
acc – accelerometer measurement [m/s^2]
mag – magnetometer measurement or None [any unit]
debug – enables debug output

Returns:

quat: orientation quaternion
debug: dict with debug values (only if debug==True)

class qmt.OriEstIMUBlock(Ts)[source]

OriEstIMU orientation estimation algorithm.

See https://dx.doi.org/10.1016/j.ifacol.2017.08.1534 for more information about this algorithm.

This algorithm is also available as a function: qmt.oriEstIMU().

Parameters:: Ts – sampling time in seconds

command(command)[source]

Supported commands:

[‘reset’]: resets the orientation estimation state

setParams(params)[source]

Supported parameters:

tauAcc: time constants for acc correction (50% time) [must be >0], [in seconds]
tauMag: time constants for mag correction (50% time) [must be >0], [in seconds]
zeta: bias estimation strength [no unit]
accRating: enables raw rating of accelerometer, set to 0 to disable

step(gyr, acc, mag, debug=False)[source]

Parameters:

gyr – gyroscope measurement [rad/s]
acc – accelerometer measurement [m/s^2]
mag – magnetometer measurement or None [any unit]
debug – enables debug output

Returns:

quat: orientation quaternion
debug: dict with debug values (only if debug==True)

class qmt.LowpassFilterBlock(Ts, fc=None, tau=None)[source]

Second-order Butterworth low-pass filter block.

This block uses the filter implementation of the VQF orientation estimation algorithm. The cutoff frequency can either directly be specified in Hz, or a time constant can be given. This time constant corresponds to the cutoff frequency as follows: \(f_\mathrm{c} = \frac{\sqrt{2}}{2\pi\tau}\).

For the first \(\tau\) seconds, the filter output is the mean of all previous samples, to ensure fast initial convergence.

Parameters:

Ts – sampling time in seconds
fc – cutoff frequency in Hz
tau – time constant in seconds (either fc or tau has to be None)

step(signal)[source]

Performs one data processing step.

The default implementation just returns None.

Parameters:

args – application-specific inputs
kwargs – application-specific inputs

Returns:

application-specific outputs