Complementary Filter Library

Introduction

This library provides a generic implementation of a first-order complementary filter for sensor fusion.

A complementary filter combines measurements from different sensors by exploiting their complementary frequency characteristics. A low-frequency measurement from one sensor can be combined with a high-frequency measurement from another sensor to obtain a more stable estimate.

The library provides a templated estimators::cf::CF class that supports different numeric types such as float and double.

Features

The complementary filter library provides:

• First-order complementary filtering.
• Fusion of two sensor measurements.
• Configurable cutoff frequency.
• Configurable sampling time.
• Configurable filter time constant.
• Configurable complementary filter coefficient.
• Runtime parameter modification.
• Filter state reset functionality.
• Support for different numeric data types through templates.

Basic Usage

A complementary filter can be created and initialized using a cutoff frequency and sampling time:

#include "cf.h"
 
estimators::cf::CF<double> filter;
 
filter.init(2.5, 0.006);
 
double output = filter.update(x1, x2);

Filter Parameters

The filter can be configured using the following parameters:

• fc : Cutoff frequency of the filter.
• tau : Time constant of the filter.
• dt : Sampling time or control-loop period.
• alpha: Complementary filter coefficient.

The parameters can be configured during initialization or individually using the corresponding setter functions.

Parameter Configuration

The filter can be initialized using:

filter.init(fc, dt);

Alternatively, individual parameters can be configured using:

filter.set_fc(fc);
filter.set_tau(tau);
filter.set_dt(dt);
filter.set_alpha(alpha);

Updating the Filter

New sensor measurements are provided to the filter using the estimators::cf::CF::update function:

double output = filter.update(x1, x2);

The two input measurements are combined according to the configured complementary filter parameters, and the filtered output is returned.

Resetting the Filter

The internal filter state can be reset using:

filter.reset();

Accessing Parameters

The current filter parameters can be obtained using:

double fc    = filter.get_fc();
double tau   = filter.get_tau();
double dt    = filter.get_dt();
double alpha = filter.get_alpha();

Main Class

The main class provided by this library is:

• estimators::cf::CF - Generic complementary filter for sensor fusion.

Namespace

The complementary filter implementation is contained in:

estimators::cf

Complete Example

#include "cf.h"
 
int main()
{
    estimators::cf::CF<double> filter;
 
    filter.init(2.5, 0.006);
 
    double x1 = 10.0;
    double x2 = 10.5;
 
    double output = filter.update(x1, x2);
 
    return 0;
}

Detailed Documentation

Detailed documentation for the complementary filter class and its member functions is available through the estimators::cf::CF class reference.

Copyright

Copyright (c) 2026 Abhinay Kumar

Author

Abhinay Kumar

Version

1.0

Date

2026-08-11