idopt

Contents Functions Index

PDF

idopt

Purpose

handling of options used by the identification related routines.

Syntax

idopt
iop2 = idopt
IDopt.OptName = OptValue;

Description

idopt should be used trough idcom(see also tutorial in section 3.3). Thus ci=idcom; IDopt=ci.IDopt displays the identification option field ci.IDopt as a detailed list of the options returns

  (ID options in figure(2)) = 
   ResidualTerms : [ 0 | 1 (1) | 2 (s^-2) | {3 (1 s^-2)} | 10 (1 s)]
   DataType : [ {disp./force} | vel./force | acc./force ]
   AbscissaUnits : [ {Hz} | rd/s | s ]
   PoleUnits : [ {Hz} | rd/s ]
   SelectedRange : [ 1-3124 (4.0039-64.9998) ]
   FittingModel : [ Posit. cpx | {Complex modes} | Normal Modes]
   NSNA : [ 0 sensor(s) 0 actuator(s) ]
   Reciprocity : [ {Not used} | 1 FRF | MIMO ]
   Collocated : [ none declared ]

with the currently selected value shown between braces { }. note that when these options are linked to a figure this is shown in the display (as above). To make a copy of the data, and no longer point to the figure, use IDopt.GetData.

iop2 = idopt returns a SDT handle to a set options that may differ from those of used by idcom.

SDT handle overloads the MATLAB getfield and setfield commands so that you can easily access each option. IDopt.OptName displays the associated option value using the format shown above. IDopt.OptName=OptValue sets the option. OptName need only specify enough characters to allow a unique option match. Thus IDopt.res and IDopt.ResidualTerms are equivalent. Typical option sets would be

 IDopt.res = 2; IDopt.sel=[1 1024]; IDopt.Po='Hz';

The following is a list of possible options with indications as to where they are stored. Thus IDopt.res=2 is simply a user friendly form for the old call IDopt(6)=2 which you can still use.

Res		Residual terms (stored in IDopt(1))
	0	none
	1	Static correction (high frequency mode correction)
	2	Roll-off (s^-2, low frequency mode correction).
	3	Static correction and roll-off (default)
	10	1 and s, this correction is only supported by id_rc and should be used for identification in narrow bandwidth (see ii_poest for example)
	-i	An alternate format uses negative numbers with decades indicating powers (starting at s^-2). Thus Ass=-1101 means an asymptotic correction with terms in s^-2,1,s
Data		type (stored in IDopt(2))
	0	displacement/force (default)
	1	velocity/force
	2	acceleration/force
Abscissa		units for vector w can be Hz, rad/s or seconds
Pole		units can be Hz or rad/s
		units are actually stored in IDopt(3) with units giving abscissa units (01 w in Hertz, 02 w in rad/s, 03 w time seconds) and tens pole units (10 po in Hertz, 20 po in rad/s). Thus IDopt (3) = 12 gives w in rad/sec and po in Hz.
Selected		frequency range indices of first and last frequencies to be used for identification or display (stored in IDopt(4:5))
Fitting		model (see res , stored in IDopt(6))
	0	positive-imaginary poles only, complex mode residue
	1	complex mode residue, pairs of complex-conjugate poles (default)
	2	normal mode residue
ns,na		number of sensors/actuators (outputs/inputs) stored in IDopt(7:8))

Recip		method selection for the treatment of reciprocity (stored in IDopt(12))
	1	means that only iC1 (IDopt(13)) is declared as being collocated. id_rm assumes that only this transfer is reciprocal even if the system has more collocated FRFs
	na	(number of actuators) is used to create fully reciprocal (and minimal of course) MIMO models using id_rm. na must match non-zero values declared in iCi.
	-nc	(with nc the number of collocated FRFs) is used to declare collocated FRFs while not enforcing reciprocity when using id_rm.
iC1 ...		indices of collocated transfer functions in the data matrix (see the xf format page )