Anda di halaman 1dari 15

1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

modalfrf
Frequency-response functions for modal analysis

Syntax

frf = modalfrf(x,y,fs,window)
frf = modalfrf(x,y,fs,window,noverlap)

frf = modalfrf( ___ ,Name,Value)

[frf,f,coh] = modalfrf( ___ )

[frf,f] = modalfrf(sys)
frf = modalfrf(sys,f)

modalfrf( ___ )

Description
frf = modalfrf(x,y,fs,window) estimates a matrix of frequency response functions, frf, from the example
excitation signals, x, and the response signals, y, all sampled at a rate fs. The output, frf, is an H estimate 1

computed using Welch’s method with window to window the signals. x and y must have the same number of
rows. If x or y is a matrix, each column represents a signal. The frequency-response function matrix, frf, is
computed in terms of dynamic flexibility, and the system response, y, contains acceleration measurements.

frf = modalfrf(x,y,fs,window,noverlap) specifies noverlap samples of overlap between adjoining


segments.

frf = modalfrf( ___ ,Name,Value) specifies options using name-value pair arguments, using any example
combination of inputs from previous syntaxes. Options include the estimator, the measurement
configuration, and the type of sensor measuring the system response.

[frf,f,coh] = modalfrf( ___ ) also returns the frequency vector corresponding to each frequency- example
response function, as well as the multiple coherence matrix.

[frf,f] = modalfrf(sys) computes the frequency-response function of the identified model sys. Use
estimation commands like ssest, n4sid, or tfest to create sys from time-domain input and output signals.
This syntax allows use only of the 'Sensor' name-value pair argument. You must have a System
Identification Toolbox™ license to use this syntax.

frf = modalfrf(sys,f) specifies the frequencies at which to compute frf. This syntax allows use only of
the 'Sensor' name-value pair argument. You must have a System Identification Toolbox license to use this
syntax.

modalfrf( ___ ) with no output arguments plots the frequency response functions in the current figure. The example
plots are limited to the first four excitations and four responses.

Examples collapse all

 Frequency-Response Function of Hammer Excitation

https://www.mathworks.com/help/signal/ref/modalfrf.html 1/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

Visualize the frequency-response function of a single-input/single-output


hammer excitation. Try it in MATLAB

Load a data file that contains:

• Xhammer An input excitation signal consisting of five hammer blows delivered periodically.
• Yhammer The response of a system to the input. Yhammer is measured as a displacement.

The signals are sampled at 4 kHz. Plot the excitation and output signals.

load modaldata

subplot(2,1,1)
plot(thammer,Xhammer(:))
ylabel('Force (N)')
subplot(2,1,2)
plot(thammer,Yhammer(:))
ylabel('Displacement (m)')
xlabel('Time (s)')

Compute and display the frequency-response function. Window the signals using a rectangular window. Specify that
the window covers the period between hammer blows.

clf
winlen = size(Xhammer,1);
modalfrf(Xhammer(:),Yhammer(:),fs,winlen,'Sensor','dis')

https://www.mathworks.com/help/signal/ref/modalfrf.html 2/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

 MIMO Frequency-Response Functions

Compute the frequency-response functions for a two-input/two-output


system excited by random noise. Try it in MATLAB

Load a data file that contains Xrand, the input excitation signal, and
Yrand, the system response. Compute the frequency-response functions using a 5000-sample Hann window and
50% overlap between adjoining data segments. Specify that the output measurements are displacements.

load modaldata
winlen = 5000;

frf = modalfrf(Xrand,Yrand,fs,hann(winlen),0.5*winlen,'Sensor','dis');

Use the plotting functionality of modalfrf to visualize the responses.

modalfrf(Xrand,Yrand,fs,hann(winlen),0.5*winlen,'Sensor','dis')

https://www.mathworks.com/help/signal/ref/modalfrf.html 3/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

 Frequency-Response Function of SISO System

Estimate the frequency-response function for a simple single-


input/single-output system and compare it to the definition. Try it in MATLAB

A one-dimensional discrete-time oscillating system consists of a unit


mass, , attached to a wall by a spring with elastic constant . A sensor samples the displacement of the mass
at Hz. A damper impedes the motion of the mass by exerting on it a force proportional to speed, with damping
constant .

Generate 3000 time samples. Define the sampling interval .

Fs = 1;
dt = 1/Fs;
N = 3000;
t = dt*(0:N-1);
b = 0.01;

The system can be described by the state-space model

https://www.mathworks.com/help/signal/ref/modalfrf.html 4/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

where is the state vector, and are respectively the displacement and velocity of the mass, is the
driving force, and is the measured output. The state-space matrices are

is the identity, and the continuous-time state-space matrices are

Ac = [0 1;-1 -b];
A = expm(Ac*dt);

Bc = [0;1];
B = Ac\(A-eye(2))*Bc;

C = [1 0];
D = 0;

The mass is driven by random input for the first 2000 seconds and then left to return to rest. Use the state-space
model to compute the time evolution of the system starting from an all-zero initial state. Plot the displacement of the
mass as a function of time.

rng default
u = randn(1,N)/2;
u(2001:end) = 0;

y = 0;
x = [0;0];
for k = 1:N
y(k) = C*x + D*u(k);
x = A*x + B*u(k);
end

plot(t,y)

https://www.mathworks.com/help/signal/ref/modalfrf.html 5/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

Estimate the modal frequency-response function of the system. Use a Hann window half as long as the measured
signals. Specify that the output is the displacement of the mass.

wind = hann(N/2);

[frf,f] = modalfrf(u',y',Fs,wind,'Sensor','dis');

The frequency-response function of a discrete-time system can be expressed as the Z-transform of the time-domain
transfer function of the system, evaluated at the unit circle. Compare the modalfrf estimate with the definition.

[b,a] = ss2tf(A,B,C,D);

nfs = 2048;
fz = 0:1/nfs:1/2-1/nfs;
z = exp(2j*pi*fz);
ztf = polyval(b,z)./polyval(a,z);

plot(f,20*log10(abs(frf)))
hold on
plot(fz*Fs,20*log10(abs(ztf)))
hold off
grid
ylim([-60 40])

https://www.mathworks.com/help/signal/ref/modalfrf.html 6/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

Estimate the natural frequency and the damping ratio for the vibration mode.

[fn,dr] = modalfit(frf,f,Fs,1,'FitMethod','PP')

fn = 0.1593
dr = 0.0043
Compare the natural frequency to , which is the theoretical value for the undamped system.

theo = 1/(2*pi)

theo = 0.1592

 Modal Parameters of Two-Body Oscillator

Estimate the frequency-response function and modal parameters of a


simple multi-input/multi-output system. Try it in MATLAB

An ideal one-dimensional oscillating system consists of two masses,


and , confined between two walls. The units are such that and . Each mass is attached to the
nearest wall by a spring with an elastic constant . An identical spring connects the two masses. Three dampers
impede the motion of the masses by exerting on them forces proportional to speed, with damping constant .
Sensors sample and , the displacements of the masses, at Hz.

https://www.mathworks.com/help/signal/ref/modalfrf.html 7/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

Generate 30000 time samples, equivalent to 600 seconds. Define the sampling interval .

Fs = 50;
dt = 1/Fs;
N = 30000;
t = dt*(0:N-1);

The system can be described by the state-space model

where is the state vector, and are respectively the location and the velocity of the th
mass, is the vector of input driving forces, and is the output vector. The state-space
matrices are

is the identity, and the continuous-time state-space matrices are

Set , , and .

k = 400;
b = 0.1;
m = 1/10;

Ac = [0 1 0 0;-2*k -2*b k b;0 0 0 1;k/m b/m -2*k/m -2*b/m];


A = expm(Ac*dt);
Bc = [0 0;1 0;0 0;0 1/m];
B = Ac\(A-eye(4))*Bc;
C = [1 0 0 0;0 0 1 0];
D = zeros(2);

The masses are driven by random input throughout the measurement. Use the state-space model to compute the
time evolution of the system starting from an all-zero initial state.

rng default
u = randn(2,N);

y = [0;0];
x = [0;0;0;0];
for kk = 1:N
y(:,kk) = C*x + D*u(:,kk);
x = A*x + B*u(:,kk);
end

Use the input and output data to estimate the transfer function of the system as a function of frequency. Use a 5000-
sample Hann window with 3000 samples of overlap between adjoining segments. Specify that the measured outputs
are displacements.

wind = hann(15000);
https://www.mathworks.com/help/signal/ref/modalfrf.html 8/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

nove = 9000;
[FRF,f] = modalfrf(u',y',Fs,wind,nove,'Sensor','dis');

Compute the theoretical transfer function as the Z-transform of the time-domain transfer function, evaluated at the unit
circle.

nfs = 2048;
fz = 0:1/nfs:1/2-1/nfs;
z = exp(2j*pi*fz);

[b1,a1] = ss2tf(A,B,C,D,1);
[b2,a2] = ss2tf(A,B,C,D,2);

frf(1,:,1) = polyval(b1(1,:),z)./polyval(a1,z);
frf(1,:,2) = polyval(b1(2,:),z)./polyval(a1,z);
frf(2,:,1) = polyval(b2(1,:),z)./polyval(a2,z);
frf(2,:,2) = polyval(b2(2,:),z)./polyval(a2,z);

Plot the estimates and overlay the theoretical predictions.

for jk = 1:2
for kj = 1:2
subplot(2,2,2*(jk-1)+kj)
plot(f,20*log10(abs(FRF(:,jk,kj))))
hold on
plot(fz*Fs,20*log10(abs(frf(jk,:,kj))))
hold off
axis([0 Fs/2 -100 0])
title(sprintf('Input %d, Output %d',jk,kj))
end
end

https://www.mathworks.com/help/signal/ref/modalfrf.html 9/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

Plot the estimates by using the syntax of modalfrf with no output arguments.

figure
modalfrf(u',y',Fs,wind,nove,'Sensor','dis')

Estimate the natural frequencies, damping ratios, and mode shapes of the system. Use the peak-picking method for
the calculation.

[fn,dr,ms] = modalfit(FRF,f,Fs,2,'FitMethod','pp');
fn

fn =
fn(:,:,1) =

3.8466 3.8466
3.8495 3.8495

fn(:,:,2) =

3.8492 3.8490
3.8552 14.4684

Compare the natural frequencies to the theoretical predictions for the undamped system.

undamped = sqrt(eig([2*k -k;-k/m 2*k/m]))/2/pi

undamped =

3.8470

https://www.mathworks.com/help/signal/ref/modalfrf.html 10/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

14.4259

 Frequency-Response Function Using Subspace Method

Compute the frequency-response function of a two-input/six-output data


set corresponding to a steel frame. Try it in MATLAB

Load a structure containing the input excitations and the output


accelerometer measurements. The system is sampled at 1024 Hz for about 3.9 seconds.

load modaldata SteelFrame


X = SteelFrame.Input;
Y = SteelFrame.Output;
fs = SteelFrame.Fs;

Use the subspace method to compute the frequency-response functions. Divide the input and output signals into
nonoverlapping, 1000-sample segments. Window each segment using a rectangular window. Specify a model order
of 36.

[frf,f] = modalfrf(X,Y,fs,1000,'Estimator','subspace','Order',36);

Visualize the stabilization diagram for the system. Identify up to 15 physical modes.

modalsd(frf,f,fs,'MaxModes',15)

Input Arguments collapse all

https://www.mathworks.com/help/signal/ref/modalfrf.html 11/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

x — Excitation signals
 vector | matrix

Excitation signals, specified as a vector or matrix.

Data Types: single | double

y — Response signals
 vector | matrix

Response signals, specified as a vector or matrix.

Data Types: single | double

fs — Sample rate
 positive scalar

Sample rate, specified as a positive scalar expressed in hertz.

Data Types: single | double

window — Window
 integer | vector

Window, specified as an integer or as a row or column vector. Use window to divide the signal into segments:

• If window is an integer, then modalfrf divides x and y into segments of length window and windows each
segment with a rectangular window of that length.
• If window is a vector, then modalfrf divides x and y into segments of the same length as the vector and windows
each segment using window.
• If 'Estimator' is specified as 'subspace', then modalfrf ignores the shape of window and uses its length to
determine the number of frequency points in the returned frequency-response function.

If the length of x and y cannot be divided exactly into an integer number of segments with noverlap overlapping
samples, then the signals are truncated accordingly.

For a list of available windows, see Windows.

Example: hann(N+1) and (1-cos(2*pi*(0:N)'/N))/2 both specify a Hann window of length N + 1.

Data Types: single | double

noverlap — Number of overlapped samples


 0 (default) | positive integer

Number of overlapped samples, specified as a positive integer.

• If window is a scalar, then noverlap must be smaller than window.


• If window is a vector, then noverlap must be smaller than the length of window.

Data Types: double | single

https://www.mathworks.com/help/signal/ref/modalfrf.html 12/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

sys — Identified system


 model with identified parameters

Identified system, specified as a model with identified parameters. Use estimation commands like ssest, n4sid, or
tfest to create sys from time-domain input and output signals. See Modal Analysis of Identified Models for an
example. Syntaxes that use sys typically require less data than syntaxes that use nonparametric methods. You must
have a System Identification Toolbox license to use this input argument.

Example: idss([0.5418 0.8373;-0.8373 0.5334],[0.4852;0.8373],[1 0],0,[0;0],[0;0],1) generates an


identified state-space model corresponding to a unit mass attached to a wall by a spring of unit elastic constant and a
damper with constant 0.01. The displacement of the mass is sampled at 1 Hz.

Example: idtf([0 0.4582 0.4566],[1 -1.0752 0.99],1) generates an identified transfer-function model
corresponding to a unit mass attached to a wall by a spring of unit elastic constant and a damper with constant 0.01.
The displacement of the mass is sampled at 1 Hz.

f — Frequencies
 vector

Frequencies, specified as a vector expressed in Hz.

Data Types: single | double

Name-Value Pair Arguments


Specify optional comma-separated pairs of Name,Value arguments. Name is the argument name and Value is the
corresponding value. Name must appear inside single quotes (' '). You can specify several name and value pair
arguments in any order as Name1,Value1,...,NameN,ValueN.

Example: 'Sensor','vel','Est','H1' specifies that the response signal consists of velocity measurements and that the
estimator of choice is H1.
collapse all

'Estimator' — Estimator
 'H1' (default) | 'H2' | 'Hv' | 'subspace'

Estimator, specified as the comma-separated pair consisting of 'Estimator' and 'H1', 'H2', 'Hv', or 'subspace'.
See Transfer Function for more information about the H and H estimators.
1 2

• Use 'H1' when the noise is uncorrelated with the excitation signals.
• Use 'H2' when the noise is uncorrelated with the response signals. In this case, the number of excitation signals
must equal the number of response signals.
• Use 'Hv' to minimize the discrepancy between modeled and estimated response data by minimizing the trace of
the error matrix. H is the geometric mean of H and H : H  = (H H )
v 1 2 v 1 2
1/2

The measurement must be single-input/single-output (SISO).


• Use 'subspace' to compute the frequency-response functions using a state-space model. In this case, the
noverlap argument is ignored. This method typically requires less data than nonparametric approaches. See
n4sid for more information.

Data Types: char

https://www.mathworks.com/help/signal/ref/modalfrf.html 13/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

'Feedthrough' — Presence of feedthrough in state-space model


 false (default) | true

Presence of feedthrough in state-space model, specified as the comma-separated pair consisting of 'Feedthrough'
and a logical value. This argument is available only if 'Estimator' is specified as 'subspace'.

Data Types: logical

'Measurement' — Measurement configuration


 'fixed' (default) | 'rovinginput' | 'rovingoutput'

Measurement configuration for equal numbers of excitation and response channels, specified as the comma-
separated pair consisting of 'Measurement' and 'fixed', 'rovinginput', or 'rovingoutput'.

• Use 'fixed' when there are excitation sources and sensors at fixed locations of the system. Each excitation
contributes to every response.
• Use 'rovinginput' when the measurements result from a roving excitation (or roving hammer) test. A single
sensor is kept at a fixed location of the system. A single excitation source is placed at multiple locations and
produces one sensor response per location. The function output frf(:,:,i) = modalfrf(x(:,i),y(:,i)).
• Use 'rovingoutput' when the measurements result from a roving sensor test. A single excitation source is kept
at a fixed location of the system. A single sensor is placed at multiple locations and responds to one excitation per
location. The function output frf(:,i) = modalfrf(x(:,i),y(:,i)).

Data Types: char

'Order' — State-space model order


 1:10 (default) | integer | row vector of integers

State-space model order, specified as the comma-separated pair consisting of 'Order' and an integer or row vector
of integers. If you specify a vector of integers, then the function selects an optimal order value from the specified
range. This argument is available only if 'Estimator' is specified as 'subspace'.

Data Types: single | double

'Sensor' — Sensor type


 'acc' (default) | 'dis' | 'vel'

Sensor type, specified as the comma-separated pair consisting of 'Sensor' and 'acc', 'dis', or 'vel'.

• 'acc' — The response signal voltage is proportional to acceleration.


• 'dis' — The response signal voltage is proportional to displacement.
• 'vel' — The response signal voltage is proportional to velocity.

Data Types: char

Output Arguments collapse all

frf — Frequency-response functions


 vector | matrix | 3-D array

https://www.mathworks.com/help/signal/ref/modalfrf.html 14/15
1/13/2018 Frequency-response functions for modal analysis - MATLAB modalfrf

Frequency-response functions, returned as a vector, matrix, or 3-D array. frf has size p-by-m-by-n, where p is the
number of frequency bins, m is the number of responses, and n is the number of excitation signals.

f — Frequencies
 vector

Frequencies, returned as a vector.

coh — Multiple coherence matrix


 matrix

Multiple coherence matrix, returned as a matrix. coh has one column for each response signal.

References
[1] Brandt, Anders. Noise and Vibration Analysis: Signal Analysis and Experimental Procedures. Chichester, UK: John
Wiley & Sons, 2011.

[2] Vold, Håvard, John Crowley, and G. Thomas Rocklin. “New Ways of Estimating Frequency Response Functions.” Sound
and Vibration. Vol. 18, November 1984, pp. 34–38.

See Also
modalfit | modalsd | n4sid | tfestimate

Topics
Modal Analysis of Identified Models
System Identification Overview (System Identification Toolbox)
System Identification Workflow (System Identification Toolbox)
Supported Continuous- and Discrete-Time Models (System Identification Toolbox)

Introduced in R2017a

https://www.mathworks.com/help/signal/ref/modalfrf.html 15/15

Anda mungkin juga menyukai