Description

Function description

This Function block enables the autotuning of the PID controller (PIDFF: Complete PID controller, PI_B: Simple PI controller).

Autotuning stabilizes the control when starting the system and, in so doing, saves time.

and can be configured as additional parameters.

Algorithm

The algorithm is based upon heuristic controls, as with the Ziegler Nichols method. Initially, an analysis corresponding to approximately 2.5 times the reaction time of the open loop is performed. Through this, the process can be identified as a process of the first order with delay.

Building on this model, a control parameter set based on heuristic controls and historical data is created.

The parameter range is determined by the ’perf’ criteria. In this individual case, this factor gives the highest rank to the reaction time to disturbances or stability.

The algorithm is applied to the following process types :

Processes with only one input / output
Processes with natural stability or integral components
Asymmetric processes within the limits authorized by the algorithm of the PID controller
Processes controlled via pulse width modulation output (PWM).

Important characteristics

The block has the following characteristics

Pre-estimation of the control for the types PIDFF and/or PI_B
Diagnostic function
Parametering of the control dynamic
Recovery of previous control settings

Operating mode

The various operating modes of the autotuning and their priorities in descending order of validity are shown in the following table:

Operating mode	TR_S	START
Tracking	1	0 or 1
Autotuning	0	1

On completion of the autotuning, the TRS output is set to 0, and the servo-loop is reset to its previous operating mode (manual or automatic). If the autotuning fails, the TRI variable will be set back to the value before autotuning was started and the servo-loop will be reset to its previous operating mode.

Representation in FBD

Representation:

Parameters of the autotuned controller (Para_PIDFF, Para_PI_B,etc.)

Representation in LD

Representation:

Parameters of the autotuned controller (Para_PIDFF, Para_PI_B,etc.)

Representation in IL

Representation:

CAL AUTOTUNE_Instance (PV:=ProcessValue, SP:=Setpoint, 
    RCPY:=RecopyRealCommand, START:=StartAutotuning, 
    PREV:=ReturnToPreviousValues, PARA:=Parameters, 
    TR_I:=InitializationInput, TR_S:=InitializationSequence, 
    PV_O=>PV_InputImage, SP_O=>SP_InputImage, 
    PARA_C=>ParametersToBeTuned, TRI=>TR_I_InputImage, 
    TRS=>TR_S_InputImage, INFO=>Information, 
    STATUS=>Statusword)

Representation in ST

Representation:

AUTOTUNE_Instance (PV:=ProcessValue, SP:=Setpoint, 
    RCPY:=RecopyRealCommand, START:=StartAutotuning, 
    PREV:=ReturnToPreviousValues, PARA:=Parameters, 
    TR_I:=InitializationInput, TR_S:=InitializationSequence, 
    PV_O=>PV_InputImage, SP_O=>SP_InputImage, 
    PARA_C=>ParametersToBeTuned, TRI=>TR_I_InputImage, 
    TRS=>TR_S_InputImage, INFO=>Information, 
    STATUS=>Statusword) ;

Parameter description AUTOTUNE

Input parameter description:

Parameter	Data type	Meaning
PV	REAL	Process value
SP	REAL	Setpoint
RCPY	REAL	Copy of the actual manipulated variable
START	BOOL	"0 → 1" : Starting the autotune
PREV	BOOL	Reverting to the previous controller settings
PARA	Para_AUTOTUNE	Parameter
TR_I	REAL	Start input
TR_S	BOOL	Start command

Output parameter description:

Parameter	Data type	Meaning
PV_O	REAL	Copy of the actual value PV
SP_O	REAL	Copy of the SP input
PARA_C	Parameters of the autotunable controller (Para_PIDFF or Para_PI_B)	Control parameters
TRI	REAL	Copy of the TR_I input
TRS	BOOL	Copy of the TR_S input
INFO	Info_AUTOTUNE	Information
STATUS	WORD	Status word

Parameter description Para_AUTOTUNE

Data structure description

Element	Data type	Meaning
step_ampl	REAL	Value of the output actuating pulse (expressed in output scale values out_inf, out_sup)
tmax	TIME	Duration of the actuating pulse in automatic tuning
perf	REAL	Performance index between 0 and 1
plant_type	WORD	Reserved word

Info_AUTOTUNE parameter description

Data structure description

Element	Data type	Meaning
diag	UDINT	Double word used for diagnosis
p1_prev	REAL	Previous value of parameter 1
p2_prev	REAL	Previous value of parameter 2
p3_prev	REAL	Previous value of parameter 3
p4_prev	REAL	Previous value of parameter 4
p5_prev	REAL	Previous value of parameter 5
p6_prev	REAL	Previous value of parameter 6

Base	Lower limit	Upper limit
Hexadecimal	16#0	16#FFFF
Octal	8#0	8#177777
Binary	2#0	2#1111111111111111

Data	Representation in one of the bases
0000000011010011	16#D3
1010101010101010	8#125252
0000000011010011	2#11010011

REAL

The REAL type is encoded in a 32 bit format.

The possible value ranges are shown in the figure below:

When a result is:

between -1,175494e-38 and 1,175494e-38, it is considered to be a DEN;
less than -3.402824e+38, the symbol -INF (for - infinity) is displayed;
greater than +3.402824e+38, the symbol INF (for + infinity) is displayed;
undefined (square root of a negative number), the symbolNAN is displayed.

NOTE: The IEC 559 standard defines two classes of NAN: the silent NAN (QNAN) and the signaling NAN (SNAN). A QNAN is a NAN with a most significant fraction bit while an SNAN is a NAN without a most significant fraction bit (bit number 22). QNANs can be propagated via most arithmetic operations without throwing an exception. As for SNANs, they generally indicate n invalid operation when they are used as operands in arithmetic operations (see %SW17 and %S18).

NOTE: When a DEN (non-standardized number) is used as an operand, the result is not significant.