pid.cpp File Reference

Implementation of all control PID systems. More...

#include "pid.h"
#include "uarm_lib.hpp"
#include "uarm_math.hpp"

Functions
void	pid_param_init (PID_t *pid, int32_t max_out, float max_i_out, float max_err, float kp, float ki, float kd)
	Initializes a PID_t structure with provided parameters.
float	pid_calculate (PID_t *pid, float cur_val, float target_val, float dt)
	Computes a single PID iteration for one time-step.
float	pid_incremental_calculate (PID_t *pid, float cur_val, float target_val)
	Calculates incremental PID output for one time-step.
float	pid_single_loop_control (float target_val, PID_t *pid, float cur_val, float dt)
	Calculates iteration of single-loop PID controller.
float	pid_dual_loop_control (float f_tar_val, PID_t *f_pid, PID_t *s_pid, float f_cur_val, float s_cur_val, float dt)
	Calculate an iteration of dual-loop cascade PID controller.
void	pid2_init (PID2_t &pid, float k_p, float k_i, float k_d, float beta, float yeta, float min_out, float max_out)
	Initializes a PID2_t controller.
void	pid2_set_limits (PID2_t &pid, float new_min_out, float new_max_out)
	Sets new output limits for PID2_t controller.
float	pid2_calculate (PID2_t &pid, float sp, float pv, float dt)
	Calculates PID2 output for one time-step.
float	pid2_single_loop_control (PID2_t &pid, float sp, float pv, float dt)
	Caculate an iteration of a single-loop PID2 controller.
float	pid2_dual_loop_control (PID2_t &f_pid, PID2_t &s_pid, float sp, float f_pv, float s_pv, float f_dt, float s_dt)
	Calculates an iteration of a dual-loop PID2 controller.
float	pid2_triple_loop_control (PID2_t &outer_pid, PID2_t &middle_pid, PID2_t &inner_pid, float sp, float outer_pv, float middle_pv, float inner_pv, float outer_dt, float middle_dt, float inner_dt)
	Calculates an iteration for a triple-loop PID2 controller.
void	prescaled_pid2_init (Prescaled_PID2_t *prescaled, uint32_t prescalar, float k_p, float k_i, float k_d, float beta, float yeta, float min_out, float max_out)
	Initializes a Prescaled_PID2_t structure with provided parameters.
void	prescaled_pid2_single_loop_control (Prescaled_PID2_t *prescaled, float sp, float pv, float dt)
	Calculate an iteration for prescaled single-loop PID2 controller.

Detailed Description

Implementation of all control PID systems.

All implementations for control systems that utilizes PID as the primary controller are included in this file. This includes single and cascade control loops as well as different implementations of PID with their own solutions to problems known about PID controllers in real systems (e.g. integator windup, derivative kick, etc.).

Author

James Fu

Date

2025-07-17

Function Documentation

pid2_calculate()

float pid2_calculate	(	PID2_t &	pid,
		float	sp,
		float	pv,
		float	dt )

Calculates PID2 output for one time-step.

Parameters

[in,out]	pid	Reference to PID2_t struct.
[in]	sp	Setpoint value.
[in]	pv	Plant value.
[in]	dt	Time step. Must be positive.

Returns

PID2 calculated output.

Warning

This PID implementation has a rudimentary anti-integrator windup. Back-calculation is not implemented.

pid2_dual_loop_control()

float pid2_dual_loop_control	(	PID2_t &	f_pid,
		PID2_t &	s_pid,
		float	sp,
		float	f_pv,
		float	s_pv,
		float	f_dt,
		float	s_dt )

Calculates an iteration of a dual-loop PID2 controller.

Parameters

[in,out]	f_pid	Reference to outer loop PID2_t struct.
[in,out]	s_pid	Reference to inner loop PID2_t struct.
[in]	sp	Setpoint value for outer loop.
[in]	f_pv	Plant value for outer loop.
[in]	s_pv	Plant value for inner loop.
[in]	f_dt	Time step for outer loop.
[in]	s_dt	Time step for inner loop.

Returns

PID2 calculated output from inner loop.

Note

Calculations are done with pid2_calculate.

pid2_init()

void pid2_init	(	PID2_t &	pid,
		float	k_p,
		float	k_i,
		float	k_d,
		float	beta,
		float	yeta,
		float	min_out,
		float	max_out )

Initializes a PID2_t controller.

Parameters

[in,out]	pid	Reference to PID2_t controller struct to initialize.
[in]	k_p	Proportional gain.
[in]	k_i	Integral gain.
[in]	k_d	Derivative gain.
[in]	beta	Proportional error scaling factor.
[in]	yeta	Derivative error scaling factor.

Note

beta and yeta are parameters for setpoint weighting, which is primarily used for negating effects of derivative kick.

Parameters

[in]	min_out	Minimum output value.
[in]	max_out	Maximum output value. Must be greater or equal to min_out.

pid2_set_limits()

void pid2_set_limits	(	PID2_t &	pid,
		float	new_min_out,
		float	new_max_out )

Sets new output limits for PID2_t controller.

Parameters

[in,out]	pid	Reference to PID2_t controller struct.
[in]	new_min_out	New minimum output value.
[in]	new_max_out	New maximum output value. Must be greater or equal to new_min_out.

Warning

Anti-integrator windup is not yet implemented.

Todo

Implement anti-integrator windup with new_min_out and new_max_out.

pid2_single_loop_control()

float pid2_single_loop_control	(	PID2_t &	pid,
		float	sp,
		float	pv,
		float	dt )

Caculate an iteration of a single-loop PID2 controller.

Parameters

[in,out]	pid	Reference to PID2_t controller struct.
[in]	sp	Setpoint value.
[in]	pv	Plant value.
[in]	dt	Time step.

Returns

PID2 calculated output.

Note

This is a wrapper that relies on pid2_calculate to do calculations.

pid2_triple_loop_control()

float pid2_triple_loop_control	(	PID2_t &	outer_pid,
		PID2_t &	middle_pid,
		PID2_t &	inner_pid,
		float	sp,
		float	outer_pv,
		float	middle_pv,
		float	inner_pv,
		float	outer_dt,
		float	middle_dt,
		float	inner_dt )

Calculates an iteration for a triple-loop PID2 controller.

Parameters

[in,out]	outer_pid	Reference to outer loop PID2_t struct.
[in,out]	middle_pid	Reference to middle loop PID2_t struct.
[in,out]	inner_pid	Reference to inner loop PID2_t struct.
[in]	sp	Setpoint value for outer loop.
[in]	outer_pv	Plant value for outer loop.
[in]	middle_pv	Plant value for middle loop.
[in]	inner_pv	Plant value for inner loop.
[in]	outer_dt	Time step for outer loop.
[in]	middle_dt	Time step for middle loop.
[in]	inner_dt	Time step for inner loop.

Returns

PID2 calculated output from inner loop.

Note

Uses pid2_calculate for each loop.

pid_calculate()

float pid_calculate	(	PID_t *	pid,
		float	cur_val,
		float	target_val,
		float	dt )

Computes a single PID iteration for one time-step.

Parameters

[in,out]	pid	Pointer to PID_t struct.
[in]	cur_val	Plant value.
[in]	target_val	Target value.
[in]	dt	Time step (seconds).

Warning

dt parameter is ignored and the time-step is always assumed to be 1.0.

Returns

PID calculated output.

Note

If error exceeds max_err, the output is zero and PID iteration calculations are skipped.

Deprecated

This PID implementation has known problems and will be removed in the future.

pid_dual_loop_control()

float pid_dual_loop_control	(	float	f_tar_val,
		PID_t *	f_pid,
		PID_t *	s_pid,
		float	f_cur_val,
		float	s_cur_val,
		float	dt )

Calculate an iteration of dual-loop cascade PID controller.

Parameters

[in]	f_tar_val	Outer loop target value.
[in,out]	f_pid	Pointer to outer loop PID_t controller struct.
[in,out]	s_pid	Pointer to inner loop PID_t controller struct.
[in]	f_cur_val	Outer loop current value.
[in]	s_cur_val	Inner loop current value.
[in]	dt	Time step for both loops.

Returns

PID calculated output from inner loop.

Deprecated

This PID implementation has known problems and will be removed in the future.

pid_incremental_calculate()

float pid_incremental_calculate	(	PID_t *	pid,
		float	cur_val,
		float	target_val )

Calculates incremental PID output for one time-step.

Parameters

[in,out]	pid	Pointer to PID_t controller struct.
[in]	cur_val	Plant value.
[in]	target_val	Target value.

Returns

PID calculated output.

Note

If error exceeds max_err, the output is zero and PID iteration is skipped.

Implements dead zone management: output below 10 is set to zero.

Deprecated

This PID implementation has known problems and will be removed in the future.

pid_param_init()

void pid_param_init	(	PID_t *	pid,
		int32_t	max_out,
		float	max_i_out,
		float	max_err,
		float	kp,
		float	ki,
		float	kd )

Initializes a PID_t structure with provided parameters.

Parameters

[out]	pid	Pointer to PID_t struct to initialize.
[in]	max_out	Maximum output magnitude. Must be non-negative.
[in]	max_i_out	Maximum integral output magnitude. Must be non-negative.
[in]	max_err	Maximum error value. Must be non-negative. Errors greater than this will skip a PID iteration.

Note

Errors greater than max_err will skip a PID iteration.

max_out, max_i_out, and max_err must be non-negative.

Parameters

[in]	kp	Proportional gain.
[in]	ki	Integral gain.
[in]	kd	Derivative gain.

Deprecated

This PID implementation has known problems and will be removed in the future.

pid_single_loop_control()

float pid_single_loop_control	(	float	target_val,
		PID_t *	pid,
		float	cur_val,
		float	dt )

Calculates iteration of single-loop PID controller.

Parameters

[in]	target_val	Target value.
[in,out]	pid	Pointer to PID_t controller struct.
[in]	cur_val	Plant value.
[in]	dt	Time step.

Returns

PID calculated output.

Note

Calls pid_calculate for computation.

Deprecated

This PID implementation has known problems and will be removed in the future.

prescaled_pid2_init()

void prescaled_pid2_init	(	Prescaled_PID2_t *	prescaled,
		uint32_t	prescalar,
		float	k_p,
		float	k_i,
		float	k_d,
		float	beta,
		float	yeta,
		float	min_out,
		float	max_out )

Initializes a Prescaled_PID2_t structure with provided parameters.

Parameters

prescaled	Pointer to Prescaled_PID2_t struct to initialize.
prescalar	Number of calls before PID2 calculation is performed.
k_p	Proportional gain.
k_i	Integral gain.
k_d	Derivative gain.
beta	Proportional error scaling factor.
yeta	Derivative error scaling factor.

Note

beta and yeta are parameters for setpoint weighting, which is primarily used for negating effects of derivative kick.

Parameters

min_out	Minimum output value.
max_out	Maximum output value.

prescaled_pid2_single_loop_control()

void prescaled_pid2_single_loop_control	(	Prescaled_PID2_t *	prescaled,
		float	sp,
		float	pv,
		float	dt )

Calculate an iteration for prescaled single-loop PID2 controller.

Parameters

prescaled	Pointer to Prescaled_PID2_t struct.
sp	Setpoint value.
pv	Plant (process) value.
dt	Time step.

Note

PID2 calculation is performed only after prescalar calls.

Uses pid2_calculate for calculations.

pid2_calculate is called with an accumulated time step cumsum_dt.