#include "pprz_chirp.h"
#include "std.h"

Include dependency graph for pprz_chirp.c:

Macros
#define	CHIRP_C1 4.0f

#define	CHIRP_C2 1.0f / (expf(CHIRP_C1) - 1)

Functions
void	chirp_init (struct chirp_t *chirp, float f0_hz, float f1_hz, float length_s, float current_time_s, bool exponential_chirp, bool fade_in)
	Allocate and initialize a new chirp struct. More...

void	chirp_reset (struct chirp_t *chirp, float current_time_s)
	Reset the time of the chirp. More...

bool	chirp_is_running (struct chirp_t *chirp, float current_time_s)
	Return if the current_time is within the chirp manoeuvre. More...

float	chirp_update (struct chirp_t *chirp, float current_time_s)
	Calculate the value at current_time_s and update the struct with current frequency and value. More...

Detailed Description

Author: Joost Meulenbeld Mathematical implementation of the chirp; Joost Meulenbeld

Mathematical implementation of the chirp A "chirp" or frequency sweep is a sine wave with in time increasing frequency, and can be used for system identification purposes. This registers a broad frequency spectrum.

Chirps can be made with frequency increasing linearly or exponentially with time. This is set using an argument in the chirp_init method. The latter one is better for system identification, according to [2].

The time length of the chirp is best put at a minimum of 4 / f_minimum such that low-frequency effects can be correctly discovered.

The fade_in argument determines if the chirp fades in with increasing amplitude at constant frequency (the lowest frequency). This makes it much easier to maintain the UAV around the trim position.

Usage example: // Initialize a chirp between 1 and 5 Hz during 2 seconds (for sys. id. use minimum of 4 * 1/f0) struct chirp_t chirp; chirp_init(&chirp, 1, 5, 2, get_current_time());

// During flight loop, add chirp value to stick input while (chirp_is_running(chirp, get_current_time())) { control[YAW] = stick_control + chirp_update(chirp, get_current_time()); }

[1] https://en.wikipedia.org/wiki/Chirp for a derivation of the linear chirp [2] Aircraft and Rotorcraft System Identification, 2nd edition by M. Tischler for a derivation of exponential chirp

Author: Alessandro Colllicelli Mathematical implementation of doublet for system identification

Definition in file pprz_chirp.c.

Macro Definition Documentation

◆ CHIRP_C1

#define CHIRP_C1 4.0f

Definition at line 29 of file pprz_chirp.c.

◆ CHIRP_C2

#define CHIRP_C2 1.0f / (expf(CHIRP_C1) - 1)

Definition at line 30 of file pprz_chirp.c.

Function Documentation

◆ chirp_init()

void chirp_init	(	struct chirp_t *	chirp,
		float	f0_hz,
		float	f1_hz,
		float	length_s,
		float	current_time_s,
		bool	exponential_chirp,
		bool	fade_in
	)

Allocate and initialize a new chirp struct.

set start_time to the current time

Parameters

f0_hz	Minimum frequency of the chirp in Hz
f1_hz	Maximum frequency of the chirp in Hz
length_s	Time interval in s (starting from start_time_s) in which the chirp should be carried out, excluding fade-in time
current_time_s	Current time in s, starting point of the chirp
exponential_chirp	If true, use exponential-time chirp, otherwise use linear-time chirp (see wikipedia)
fade_in	If true, begin the chirp with 2 wavelengths of the lowest frequency, with increasing amplitude