Template Class LowPassFilter

Class Documentation

template<typename T>
class LowPassFilter

A Low-pass filter class.

This class implements a low-pass filter for various data types based on an Infinite Impulse Response Filter. For vector elements, the filtering is applied separately on each element of the vector.

In particular, this class implements a simplified version of an IIR filter equation :

\(y(n) = b x(n-1) + a y(n-1)\)

where:

  • \( x(n)\) is the input signal

  • \( y(n)\) is the output signal (filtered)

  • \( b \) is the feedforward filter coefficient

  • \( a \) is the feedback filter coefficient

and the Low-Pass coefficient equation:

  • \( a = e^{\frac{-1}{sf} \frac{2\pi df}{10^{\frac{di}{-10}}}} \)

  • \( b = 1 - a \)

where:

  • \( sf \) is the sampling frequency

  • \( df \) is the damping frequency

  • \( di \) is the damping intensity (amplitude)

Usage

For manual instantiation, you should first call configure() (in non-realtime) and then call update() at every update step.

Public Functions

LowPassFilter()
inline LowPassFilter(double sampling_frequency, double damping_frequency, double damping_intensity)
~LowPassFilter()

Destructor of LowPassFilter class.

bool configure()

Configure the LowPassFilter (access and process params).

bool update(const T &data_in, T &data_out)

Applies one iteration of the IIR filter.

Parameters:

  • data_in – input to the filter

  • data_out – filtered output

Returns:

false if filter is not configured, true otherwise

inline bool is_configured() const
inline void set_params(double sampling_frequency, double damping_frequency, double damping_intensity)

Internal computation of the feedforward and feedbackward coefficients according to the LowPassFilter parameters.

inline bool update(const geometry_msgs::msg::WrenchStamped &data_in, geometry_msgs::msg::WrenchStamped &data_out)
inline bool update(const std::vector<double> &data_in, std::vector<double> &data_out)