#include <butterworth_filter.h>

Public Member Functions
	ButterworthFilter (double low_pass_filter_coeff)

	ButterworthFilter ()=delete

double	filter (double new_measurement)

void	reset (const double data)

Detailed Description

Class ButterworthFilter - Implementation of a signal filter to soften jerks. This is a first-order Butterworth low-pass filter. First-order was chosen for 2 reasons:

It doesn't overshoot
Computational efficiency This filter has been parameterized so there is only one parameter to tune. See "Digital Implementation of Butterworth First–Order Filter Type IIR" by Horvath, Cervenanska, and Kotianova, 2019 and Mienkina, M., Filter-Based Algorithm for Metering Applications, https://www.nxp.com/docs/en/application-note/AN4265.pdf, 2016 It comes from finding the bilinear transform equivalent of the analog transfer function and further applying the inverse z-transform. The parameter "low_pass_filter_coeff" equals (2*pi / tan(omega_d * T)) where omega_d is the cutoff frequency and T is the sampling period in sec.

Definition at line 65 of file butterworth_filter.h.

Constructor & Destructor Documentation

◆ ButterworthFilter() [1/2]

online_signal_smoothing::ButterworthFilter::ButterworthFilter ( double low_pass_filter_coeff )

Constructor.

Parameters

low_pass_filter_coeff Larger filter_coeff-> more smoothing of commands, but more lag. low_pass_filter_coeff = (2*pi / tan(omega_d * T)) where omega_d is the cutoff frequency and T is the sampling period in sec.

Definition at line 53 of file butterworth_filter.cpp.

◆ ButterworthFilter() [2/2]

online_signal_smoothing::ButterworthFilter::ButterworthFilter ( )

delete

Member Function Documentation

◆ filter()

double online_signal_smoothing::ButterworthFilter::filter ( double new_measurement )

Definition at line 82 of file butterworth_filter.cpp.

Here is the caller graph for this function:

◆ reset()

void online_signal_smoothing::ButterworthFilter::reset ( const double data )

Definition at line 94 of file butterworth_filter.cpp.