A mrac controller for actuation system (throttle/brake and steering) More...

#include <mrac_controller.h>

Collaboration diagram for apollo::control::MracController:

Public Member Functions
void	Init (const MracConf &mrac_conf, const common::LatencyParam &latency_param, const double dt)
	initialize mrac controller More...

common::Status	SetReferenceModel (const MracConf &mrac_conf)

common::Status	SetAdaptionModel (const MracConf &mrac_conf)

common::Status	BuildReferenceModel ()
	build mrac (1st or 2nd) order reference model in the discrete-time form, with the bilinear transform (trapezoidal integration) method More...

common::Status	BuildAdaptionModel ()
	build mrac (1st or 2nd) order adaptive dynamic model in the discrete-time form More...

bool	CheckLyapunovPD (const Eigen::MatrixXd matrix_a, const Eigen::MatrixXd matrix_p) const
	check if the solution of the algebraic Lyapunov Equation is symmetric positive definite More...

void	EstimateInitialGains (const common::LatencyParam &latency_param)
	estimate the initial states of the adaptive gains via known actuation dynamics approximation More...

void	UpdateReference ()
	execute the reference state interation with respect to the designed inputs in discrete-time form, with the bilinear transform (trapezoidal integration) method More...

void	UpdateAdaption (Eigen::MatrixXd *law_adp, const Eigen::MatrixXd state_adp, const Eigen::MatrixXd gain_adp)
	execute the adaption interation with respect to the designed law in discrete-time form, with the bilinear transform (trapezoidal integration) method More...

void	AntiWindupCompensation (const double control_command, const double previous_command)
	calculate the anti-windup compensation with respect to the integral windup issue More...

void	Reset ()
	reset all the variables (including all the states, gains and externally-setting control parameters) for mrac controller More...

void	ResetStates ()
	reset internal states for mrac controller More...

void	ResetGains ()
	reset adaptive gains for mrac controller More...

virtual double	Control (const double command, const Eigen::MatrixXd state, const double input_limit, const double input_rate_limit)
	compute control value based on the original command More...

int	BoundOutput (const double output_unbounded, const double previous_output, double *output_bounded)
	bound the system output with the given bound and change-rate bound More...

void	SetInitialReferenceState (const Eigen::MatrixXd &state_reference_init)
	set initial values for state components in reference model dynamics More...

void	SetInitialActionState (const Eigen::MatrixXd &state_action_init)
	set initial values for state components in actual actuator dynamics More...

void	SetInitialCommand (const double command_init)
	set initial command (desired input) More...

void	SetInitialStateAdaptionGain (const Eigen::MatrixXd &gain_state_adaption_init)
	set initial values of state adaption gains for mrac control More...

void	SetInitialInputAdaptionGain (const double gain_input_adaption_init)
	set initial value of input adaption gain for mrac control More...

void	SetInitialNonlinearAdaptionGain (const double gain_nonlinear_adaption_init)
	set initial value of nonlinear adaption gain for mrac control More...

void	SetStateAdaptionRate (const double ratio_state)
	set convergence ratio for state components in adaptive dynamics More...

void	SetInputAdaptionRate (const double ratio_input)
	set convergence ratio for input components in adaptive dynamics More...

void	SetNonlinearAdaptionRate (const double ratio_nonlinear)
	set convergence ratio for nonlinear components in adaptive dynamics More...

double	StateAdaptionRate () const
	get convergence ratio for state components in adaptive dynamics More...

double	InputAdaptionRate () const
	get convergence ratio for input components in adaptive dynamics More...

double	NonlinearAdaptionRate () const
	get convergence ratio for nonlinear components in adaptive dynamics More...

int	ReferenceSaturationStatus () const
	get saturation status for reference system More...

int	ControlSaturationStatus () const
	get saturation status for control system More...

Eigen::MatrixXd	CurrentReferenceState () const
	get current state for reference system More...

Eigen::MatrixXd	CurrentStateAdaptionGain () const
	get current state adaptive gain for mrac control More...

Eigen::MatrixXd	CurrentInputAdaptionGain () const
	get current input adaptive gain for mrac control More...

Eigen::MatrixXd	CurrentNonlinearAdaptionGain () const
	get current nonlinear adaptive gain for mrac control More...

Protected Attributes
bool	reference_model_enabled_ = false

bool	adaption_model_enabled_ = false

bool	adaption_clamping_enabled = false

int	model_order_ = 1

double	tau_reference_ = 0.0

double	tau_clamping_ = 0.0

double	wn_reference_ = 0.0

double	zeta_reference_ = 0.0

double	ts_ = 0.01

Eigen::MatrixXd	gamma_state_adaption_

Eigen::MatrixXd	gamma_input_adaption_

Eigen::MatrixXd	gamma_nonlinear_adaption_

double	gamma_ratio_state_ = 1.0

double	gamma_ratio_input_ = 1.0

double	gamma_ratio_nonlinear_ = 1.0

Eigen::MatrixXd	matrix_a_reference_

Eigen::MatrixXd	matrix_b_reference_

Eigen::MatrixXd	matrix_p_adaption_

Eigen::MatrixXd	matrix_b_adaption_

Eigen::MatrixXd	input_desired_

Eigen::MatrixXd	state_action_

Eigen::MatrixXd	state_reference_

Eigen::MatrixXd	gain_state_adaption_

Eigen::MatrixXd	gain_input_adaption_

Eigen::MatrixXd	gain_nonlinear_adaption_

Eigen::MatrixXd	gain_state_clamping_

Eigen::MatrixXd	gain_input_clamping_

Eigen::MatrixXd	gain_nonlinear_clamping_

Eigen::MatrixXd	gain_state_adaption_init_

Eigen::MatrixXd	gain_input_adaption_init_

Eigen::MatrixXd	gain_nonlinear_adaption_init_

double	control_previous_ = 0.0

double	bound_ratio_ = 0.0

double	bound_command_ = 0.0

double	bound_command_rate_ = 0.0

int	saturation_status_reference_ = 0

int	saturation_status_control_ = 0

Eigen::MatrixXd	gain_anti_windup_

Eigen::MatrixXd	compensation_anti_windup_

Detailed Description

A mrac controller for actuation system (throttle/brake and steering)

Member Function Documentation

◆ AntiWindupCompensation()

void apollo::control::MracController::AntiWindupCompensation	(	const double	control_command,
		const double	previous_command
	)

calculate the anti-windup compensation with respect to the integral windup issue

Parameters

control_command	desired control command for the actuator
previous_command	last control command for the actuator
dt	control sampling time

◆ BoundOutput()

int apollo::control::MracController::BoundOutput	(	const double	output_unbounded,
		const double	previous_output,
		double *	output_bounded
	)

bound the system output with the given bound and change-rate bound

Parameters

output_unbounded	original system output without bound
previous_output	system output in the last step
dt	sampling time interval
output_bounded	bounded system output

Returns: saturation_status system saturation status indicator

◆ BuildAdaptionModel()

common::Status apollo::control::MracController::BuildAdaptionModel ( )

build mrac (1st or 2nd) order adaptive dynamic model in the discrete-time form

Returns: Status adaption model initialization status

◆ BuildReferenceModel()

common::Status apollo::control::MracController::BuildReferenceModel ( )

build mrac (1st or 2nd) order reference model in the discrete-time form, with the bilinear transform (trapezoidal integration) method

Returns: Status reference model initialization status

◆ CheckLyapunovPD()

bool apollo::control::MracController::CheckLyapunovPD	(	const Eigen::MatrixXd	matrix_a,
		const Eigen::MatrixXd	matrix_p
	)		const

check if the solution of the algebraic Lyapunov Equation is symmetric positive definite

Parameters

matrix_a	reference model matrix
matrix_p	Lyapunov function matrix

Returns: indicator of the symmetric positive definite matrix

◆ Control()

virtual double apollo::control::MracController::Control	(	const double	command,
		const Eigen::MatrixXd	state,
		const double	input_limit,
		const double	input_rate_limit
	)

virtual

compute control value based on the original command

Parameters

command	original command as the input of the actuation system
state	actual output state of the actuation system
dt	sampling time interval
input_limit	physical or designed bound of the input
input_rate_limit	physical or designed bound of the input changing-rate

Returns: control value based on mrac controller architecture

◆ ControlSaturationStatus()

int apollo::control::MracController::ControlSaturationStatus ( ) const

get saturation status for control system

Returns: saturation status

◆ CurrentInputAdaptionGain()

Eigen::MatrixXd apollo::control::MracController::CurrentInputAdaptionGain ( ) const

get current input adaptive gain for mrac control

Returns: current input adaptive gain

◆ CurrentNonlinearAdaptionGain()

Eigen::MatrixXd apollo::control::MracController::CurrentNonlinearAdaptionGain ( ) const

get current nonlinear adaptive gain for mrac control

Returns: current nonlinear adaptive gain

◆ CurrentReferenceState()

Eigen::MatrixXd apollo::control::MracController::CurrentReferenceState ( ) const

get current state for reference system

Returns: current state

◆ CurrentStateAdaptionGain()

Eigen::MatrixXd apollo::control::MracController::CurrentStateAdaptionGain ( ) const

get current state adaptive gain for mrac control

Returns: current state adaptive gain

◆ EstimateInitialGains()

void apollo::control::MracController::EstimateInitialGains ( const common::LatencyParam & latency_param )

estimate the initial states of the adaptive gains via known actuation dynamics approximation

Parameters

latency_param configuration for actuation latency parameters

◆ Init()

void apollo::control::MracController::Init	(	const MracConf &	mrac_conf,
		const common::LatencyParam &	latency_param,
		const double	dt
	)

initialize mrac controller

Parameters

mrac_conf	configuration for mrac controller
latency_param	configuration for latency parameter
dt	sampling time interval

◆ InputAdaptionRate()

double apollo::control::MracController::InputAdaptionRate ( ) const

get convergence ratio for input components in adaptive dynamics

Returns: input adaption ratio

◆ NonlinearAdaptionRate()

double apollo::control::MracController::NonlinearAdaptionRate ( ) const

get convergence ratio for nonlinear components in adaptive dynamics

Returns: nonlinear adaption ratio

◆ ReferenceSaturationStatus()

int apollo::control::MracController::ReferenceSaturationStatus ( ) const

get saturation status for reference system

Returns: saturation status

◆ Reset()

void apollo::control::MracController::Reset ( )

reset all the variables (including all the states, gains and externally-setting control parameters) for mrac controller

◆ ResetGains()

void apollo::control::MracController::ResetGains ( )

reset adaptive gains for mrac controller

◆ ResetStates()

void apollo::control::MracController::ResetStates ( )

reset internal states for mrac controller

◆ SetAdaptionModel()

common::Status apollo::control::MracController::SetAdaptionModel ( const MracConf & mrac_conf )

state adaptive gain, desired adaptive gain and nonlinear-component adaptive gain

Parameters

mrac_conf configuration for adaption model

Returns: Status parameter initialization status

◆ SetInitialActionState()

void apollo::control::MracController::SetInitialActionState ( const Eigen::MatrixXd & state_action_init )

set initial values for state components in actual actuator dynamics

Parameters

state_reference_init initial action states

◆ SetInitialCommand()

void apollo::control::MracController::SetInitialCommand ( const double command_init )

set initial command (desired input)

Parameters

command_init initial desired input

◆ SetInitialInputAdaptionGain()

void apollo::control::MracController::SetInitialInputAdaptionGain ( const double gain_input_adaption_init )

set initial value of input adaption gain for mrac control

Parameters

gain_input_adaption_init initial input adaption gain

◆ SetInitialNonlinearAdaptionGain()

void apollo::control::MracController::SetInitialNonlinearAdaptionGain ( const double gain_nonlinear_adaption_init )

set initial value of nonlinear adaption gain for mrac control

Parameters

gain_nonlinear_adaption_init initial nonlinear adaption gain

◆ SetInitialReferenceState()

void apollo::control::MracController::SetInitialReferenceState ( const Eigen::MatrixXd & state_reference_init )

set initial values for state components in reference model dynamics

Parameters

state_reference_init initial reference states

◆ SetInitialStateAdaptionGain()

void apollo::control::MracController::SetInitialStateAdaptionGain ( const Eigen::MatrixXd & gain_state_adaption_init )

set initial values of state adaption gains for mrac control

Parameters

gain_state_adaption_init initial state adaption gains

◆ SetInputAdaptionRate()

void apollo::control::MracController::SetInputAdaptionRate ( const double ratio_input )

set convergence ratio for input components in adaptive dynamics

Parameters

ratio_input convergence ratio for input adaption

◆ SetNonlinearAdaptionRate()

void apollo::control::MracController::SetNonlinearAdaptionRate ( const double ratio_nonlinear )

set convergence ratio for nonlinear components in adaptive dynamics

Parameters

ratio_nonlinear convergence ratio for additional nonlinear adaption

◆ SetReferenceModel()

common::Status apollo::control::MracController::SetReferenceModel ( const MracConf & mrac_conf )

time constant, natural frequency and damping ratio

Parameters

mrac_conf configuration for reference model

Returns: Status parameter initialization status

◆ SetStateAdaptionRate()

void apollo::control::MracController::SetStateAdaptionRate ( const double ratio_state )

set convergence ratio for state components in adaptive dynamics

Parameters

ratio_state convergence ratio for state adaption

◆ StateAdaptionRate()

double apollo::control::MracController::StateAdaptionRate ( ) const

get convergence ratio for state components in adaptive dynamics

Returns: state adaption ratio

◆ UpdateAdaption()

void apollo::control::MracController::UpdateAdaption	(	Eigen::MatrixXd *	law_adp,
		const Eigen::MatrixXd	state_adp,
		const Eigen::MatrixXd	gain_adp
	)

execute the adaption interation with respect to the designed law in discrete-time form, with the bilinear transform (trapezoidal integration) method

Parameters

law_adp	adaptive law at k and k-1 steps
state_adp	state used in the adaptive law at k and k-1 steps
gain_adp	adaptive gain for the given adaptive law

◆ UpdateReference()

void apollo::control::MracController::UpdateReference ( )

execute the reference state interation with respect to the designed inputs in discrete-time form, with the bilinear transform (trapezoidal integration) method

Member Data Documentation

◆ adaption_clamping_enabled

bool apollo::control::MracController::adaption_clamping_enabled = false

protected

◆ adaption_model_enabled_

bool apollo::control::MracController::adaption_model_enabled_ = false

protected

◆ bound_command_

double apollo::control::MracController::bound_command_ = 0.0

protected

◆ bound_command_rate_

double apollo::control::MracController::bound_command_rate_ = 0.0

protected

◆ bound_ratio_

double apollo::control::MracController::bound_ratio_ = 0.0

protected

◆ compensation_anti_windup_

Eigen::MatrixXd apollo::control::MracController::compensation_anti_windup_

protected

◆ control_previous_

double apollo::control::MracController::control_previous_ = 0.0

protected

◆ gain_anti_windup_

Eigen::MatrixXd apollo::control::MracController::gain_anti_windup_

protected

◆ gain_input_adaption_

Eigen::MatrixXd apollo::control::MracController::gain_input_adaption_

protected

◆ gain_input_adaption_init_

Eigen::MatrixXd apollo::control::MracController::gain_input_adaption_init_

protected

◆ gain_input_clamping_

Eigen::MatrixXd apollo::control::MracController::gain_input_clamping_

protected

◆ gain_nonlinear_adaption_

Eigen::MatrixXd apollo::control::MracController::gain_nonlinear_adaption_

protected

◆ gain_nonlinear_adaption_init_

Eigen::MatrixXd apollo::control::MracController::gain_nonlinear_adaption_init_

protected

◆ gain_nonlinear_clamping_

Eigen::MatrixXd apollo::control::MracController::gain_nonlinear_clamping_

protected

◆ gain_state_adaption_

Eigen::MatrixXd apollo::control::MracController::gain_state_adaption_

protected

◆ gain_state_adaption_init_

Eigen::MatrixXd apollo::control::MracController::gain_state_adaption_init_

protected

◆ gain_state_clamping_

Eigen::MatrixXd apollo::control::MracController::gain_state_clamping_

protected

◆ gamma_input_adaption_

Eigen::MatrixXd apollo::control::MracController::gamma_input_adaption_

protected

◆ gamma_nonlinear_adaption_

Eigen::MatrixXd apollo::control::MracController::gamma_nonlinear_adaption_

protected

◆ gamma_ratio_input_

double apollo::control::MracController::gamma_ratio_input_ = 1.0

protected

◆ gamma_ratio_nonlinear_

double apollo::control::MracController::gamma_ratio_nonlinear_ = 1.0

protected

◆ gamma_ratio_state_

double apollo::control::MracController::gamma_ratio_state_ = 1.0

protected

◆ gamma_state_adaption_

Eigen::MatrixXd apollo::control::MracController::gamma_state_adaption_

protected

◆ input_desired_

Eigen::MatrixXd apollo::control::MracController::input_desired_

protected

◆ matrix_a_reference_

Eigen::MatrixXd apollo::control::MracController::matrix_a_reference_

protected

◆ matrix_b_adaption_

Eigen::MatrixXd apollo::control::MracController::matrix_b_adaption_

protected

◆ matrix_b_reference_

Eigen::MatrixXd apollo::control::MracController::matrix_b_reference_

protected

◆ matrix_p_adaption_

Eigen::MatrixXd apollo::control::MracController::matrix_p_adaption_

protected

◆ model_order_

int apollo::control::MracController::model_order_ = 1

protected

◆ reference_model_enabled_

bool apollo::control::MracController::reference_model_enabled_ = false

protected

◆ saturation_status_control_

int apollo::control::MracController::saturation_status_control_ = 0

protected

◆ saturation_status_reference_

int apollo::control::MracController::saturation_status_reference_ = 0

protected

◆ state_action_

Eigen::MatrixXd apollo::control::MracController::state_action_

protected

◆ state_reference_

Eigen::MatrixXd apollo::control::MracController::state_reference_

protected

◆ tau_clamping_

double apollo::control::MracController::tau_clamping_ = 0.0

protected

◆ tau_reference_

double apollo::control::MracController::tau_reference_ = 0.0

protected

◆ ts_

double apollo::control::MracController::ts_ = 0.01

protected

◆ wn_reference_

double apollo::control::MracController::wn_reference_ = 0.0

protected

◆ zeta_reference_

double apollo::control::MracController::zeta_reference_ = 0.0

protected

The documentation for this class was generated from the following file:

modules/control/common/mrac_controller.h

Public Member Functions

Protected Attributes

Detailed Description

Member Function Documentation

◆ AntiWindupCompensation()

◆ BoundOutput()

◆ BuildAdaptionModel()

◆ BuildReferenceModel()

◆ CheckLyapunovPD()

◆ Control()

◆ ControlSaturationStatus()

◆ CurrentInputAdaptionGain()

◆ CurrentNonlinearAdaptionGain()

◆ CurrentReferenceState()

◆ CurrentStateAdaptionGain()

◆ EstimateInitialGains()

◆ Init()

◆ InputAdaptionRate()

◆ NonlinearAdaptionRate()

◆ ReferenceSaturationStatus()

◆ Reset()

◆ ResetGains()

◆ ResetStates()

◆ SetAdaptionModel()

◆ SetInitialActionState()

◆ SetInitialCommand()

◆ SetInitialInputAdaptionGain()

◆ SetInitialNonlinearAdaptionGain()

◆ SetInitialReferenceState()

◆ SetInitialStateAdaptionGain()

◆ SetInputAdaptionRate()

◆ SetNonlinearAdaptionRate()

◆ SetReferenceModel()

◆ SetStateAdaptionRate()

◆ StateAdaptionRate()

◆ UpdateAdaption()

◆ UpdateReference()

Member Data Documentation

◆ adaption_clamping_enabled

◆ adaption_model_enabled_

◆ bound_command_

◆ bound_command_rate_

◆ bound_ratio_

◆ compensation_anti_windup_

◆ control_previous_

◆ gain_anti_windup_

◆ gain_input_adaption_

◆ gain_input_adaption_init_

◆ gain_input_clamping_

◆ gain_nonlinear_adaption_

◆ gain_nonlinear_adaption_init_

◆ gain_nonlinear_clamping_

◆ gain_state_adaption_

◆ gain_state_adaption_init_

◆ gain_state_clamping_

◆ gamma_input_adaption_

◆ gamma_nonlinear_adaption_

◆ gamma_ratio_input_

◆ gamma_ratio_nonlinear_

◆ gamma_ratio_state_

◆ gamma_state_adaption_

◆ input_desired_

◆ matrix_a_reference_

◆ matrix_b_adaption_

◆ matrix_b_reference_

◆ matrix_p_adaption_

◆ model_order_

◆ reference_model_enabled_

◆ saturation_status_control_

◆ saturation_status_reference_

◆ state_action_

◆ state_reference_

◆ tau_clamping_

◆ tau_reference_

◆ ts_

◆ wn_reference_

◆ zeta_reference_