0% found this document useful (0 votes)
65 views21 pages

ME 34000 and ECE 34000 Instrumentation and Measurement Systems

This document provides an overview of system identification methods for characterizing dynamic systems using measurements. It discusses identifying the parameters of first-order and second-order systems through time and frequency domain analysis. For first-order systems, the time constant can be determined from the step response or frequency response. For second-order systems, the natural frequency and damping ratio can be found from the snap-back response, step response, or frequency response. Frequency response testing involves measuring the magnitude and phase shift of a system's response to sinusoidal inputs.

Uploaded by

Mamadou Dem
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
65 views21 pages

ME 34000 and ECE 34000 Instrumentation and Measurement Systems

This document provides an overview of system identification methods for characterizing dynamic systems using measurements. It discusses identifying the parameters of first-order and second-order systems through time and frequency domain analysis. For first-order systems, the time constant can be determined from the step response or frequency response. For second-order systems, the natural frequency and damping ratio can be found from the snap-back response, step response, or frequency response. Frequency response testing involves measuring the magnitude and phase shift of a system's response to sinusoidal inputs.

Uploaded by

Mamadou Dem
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 21

ME 34000 and ECE 34000

Instrumentation and Measurement Systems


Lecture Notes: 6
Instructor
Peter Orono, Ph.D.
Dept. of Mechanical Engineering
IUPUI

System Identification

System identification is a mathematical process


of characterizing a system using measurements
obtained from the given system.
Why do we need to characterize a system?
Dynamic systems are normally represented by a
system of differential equations which contains
various system parameters such as time
constant (1st order system), natural frequency &
damping ratio (2nd order system), and so on.
Unless these parameters are accurately
determined, we may not be able to specify the
accuracy and precision of a measurement
system component.
2

System Identification

A measurement system involving acceleration


measurement can have three different components as
shown.
Each of these components have its own dynamic
characteristics (time constant, natural frequency,
damping ratio, etc.).
The accuracy and precision of acceleration measurement
depends on each of these dynamic characteristics.
Chart Recorder

Accelerometer

Amplifier

Chart Recorder

System Identification Methods

Well refer system identification as SysId for short.


SysId is a subject by itself. In depth coverage is
outside the scope of this course.
Well consider system ID methods for 1st order and
2nd order system here.
The first order system is represented by the
following equation:

y (t ) y (t ) K * F (t )

Examples of first order system are: RC or RL circuits,


thermal systems, mass air flow dynamics in an
engine, etc.
4

Response of First Order Systems

The solution to 1st order ODE is:


y(t) = yf +

0.63

SysId for 1st Order System

If we can measure the time at 63.2 percent


point of the output for a step input, we should
be able to determine the time constant of the
system.
The time constant is given by:

y (t ) y0
1 e
y f y0

It is easy to calculate the time constant from


the above equation.

SysId for 1st Order System

Another way to find the time constant is to


rearrange the equation as follows:
e

y f y0
y (t ) y f
t

ln
ym
y0 y f

y f y (t )
Slope =
-1

ym

Time, sec.

It is easy to calculate the time constant from


the slope of the above plot.
7

SysId for 1st Order System: Frequency


Response Method

Frequency response (measurement system response to


sinusoidal inputs) results can also be used to determine
the time constant of a measurement system.
How can you tell from a frequency plot (magnitude and
phase) that the system is of 1st order?
The static sensitivity K and time constant can be
estimated from these plots.
Roll-off slope:
1st order system: -20 dB/decade
2nd order system: -40 dB/decade

M(f)

SysId for 1st Order System: Frequency


Response Method

At =0 rad/sec, the magnitude of frequency response


function is |K| or 20log10|K|.
At = 1/ rad/sec, the magnitude is K/sqrt(2) or 3dB down
from 20log10|K|.
Find the frequency at which magnitude drops by 3 dB
from the flat portion (1st Order system) and then the time
constant is inverse of this frequency.

y (t ) y (t ) Kx(t )

K = |M()|

=0

3 dB

M(f)

1/

System Identification 2nd Order


system

A second order system is characterized by a


natural frequency and a damping ratio. System
identification of a second order system involves
determination of these two parameters.
A 2nd order system is represented by a 2nd order
differential equation.
1
2

y (t )
y (t ) y (t ) K * F (t )
2
n
n

10

SysID of 2nd Order System


A solution to the 2nd order ODE with non-zero
initial condition but without any forcing function
is given by:

y (t ) y0e n t cos n 1 2 t y0e n t cos d t

The snap-back response shown in the above


equation is shown below:

11

SysID of 2nd Order System

The period Td and the damped natural


frequency d are given by:

2
Td
d

d n 1

Knowledge of Td will yield d.


How do we find ?
12

SysID of 2nd Order System

It can be noted from the figure that the


envelope of the response is given by:

Envelope y0 e

Recall the snap back response:

y (t ) y0 e

n t

n t

cos d t

If you consider the peaks of the snap back


response, these points will lie on the envelope.
Lets consider two points in time where the
peaks occur: ti and ti+nTd. The second peak
occurs after n cycles.
13

SysID of 2nd Order System

Considering the general case where the final steady state


value is non-zero (yf), the response values at these two
n ti
peak points are given by:

yi y f y 0 e

yi n y f y0 e n ( ti nTd )

Taking logarithm on the above equation, one can obtain


the following:

ln( yi y f ) ln y0 nti

ln( yi n y f ) ln y0 n (ti nTd )


ln( yi y f ) ln( yi n y f ) n nTd
( yi y f )
1
nTd ln

n ( yi n y f )
14

SysID of 2nd Order System


in the above equation is known as the logarithmic
decrement that quantifies how much the response decays
over time. It can be computed from a plot of data for a
step, impulse, or snap back response. The damping ratio
is the calculated from the following
equation:
2
2

nTd n

1 2

(4 2 2 )

Once is known, the following equation can used to


obtain n:

n
15

Td 1 2

SysID of 2nd Order System

Lets now consider a step response of a second order


system. The overshoot of the step response is given by:
(from ME 330 or ECE 301)

SettlingTi me, Ts

Max OS

RiseTime , Tr

Maximum _ Overshoot e
16

4
;2%definition
n

2.16 0.6
;0.3 0.8
n

SysID of 2nd Order System

By measuring the maximum overshoot of a step


response, it is possible to determine the damping ratio of
a 2nd order system.
The natural frequency is obtained in a similar way as for
the snap back response.

Td

2
d
n 1 2
Td
17

2nd Order System ID: Frequency


Response Method

Frequency response of a 2nd order dynamic system can


also be used to determine the system parameters.
You have to give sinusoidal input to the system with
varying frequency and measure the magnitude and phase
shift of the response and plot them.

-90

18

2nd Order System ID: Frequency


Response Method

Note that a second order system will have a -40


dB/decade roll-off slope in the magnitude plot.
The natural frequency can be computed be locating the
frequency on the phase plot at which the phase lag is 90 0.
You can calculate the damping ratio using one of the
following equations.
Amplification Method

M ( )

M ( ) 0

or
Slope of Phase Angle Method

Half Power Method


19

1

n

d n

2 2

2n
2n

2nd Order System ID: Frequency


Response Method

20

Next class

Analog Measurement Devices

21

You might also like