Scribd
Upload
75 views2 pages

Recursive Least Square

1. The document simulates different inputs to a process model and observes the corresponding outputs over time. It generates step, noise, and filtered inputs and graphs the inputs and outputs. 2. It then uses recursive least squares (RLS) to estimate the process model from the data, starting the RLS both with and without prior information. It graphs the parameter estimates over time and the error between the real and estimated outputs. 3. A measurement noise signal is then added to the process. The same simulations and RLS estimations are run using the noisy data and results are compared to the noise-free case.

Uploaded by

Zinaida Latić
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
75 views2 pages

Recursive Least Square

1. The document simulates different inputs to a process model and observes the corresponding outputs over time. It generates step, noise, and filtered inputs and graphs the inputs and outputs. 2. It then uses recursive least squares (RLS) to estimate the process model from the data, starting the RLS both with and without prior information. It graphs the parameter estimates over time and the error between the real and estimated outputs. 3. A measurement noise signal is then added to the process. The same simulations and RLS estimations are run using the noisy data and results are compared to the noise-free case.

Uploaded by

Zinaida Latić
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

1.

Real process with transfer function:


1
G0 ( s ) .
( s 0.5 s 1)
2

Generate input and output of process in time 0 t 100 s , with step T=0.1 [s], for three different
cases:

a) Step signal
b) White noise signal which takes values from 0 to 2.
c) Output of filter which transfer fcn is
1
G f ( s) , if input is signal from b.
0.1s 1

2. Input and output signals from 1 show graphicly.

3. Using RLS and data generated in 1a,1b,1c estimate values of process model.
a) RLS start on two different ways.
a1) RLS starts without any informations about parametars
a2) RLS starts in t=10 [s], depending on LS estimation until t=10 [s].
b) Graphicly show estipation of parametars in time 0 t 100 s .
c) Graphicly show error between real process and estimated values on output of process in time
0 t 100 s .
d) Compare results.

4. Real process with transfer fcn:


1
G0 ( s ) .
s( s 0.5 s 1)
2

Do all steps like in 1,2 and 3.

5. For the analysis of process modeling in the conditions of measurement interference (or
inaccuracy of measurement) assume the situation as in Fig.1. where e (t) denotes the measurement
interference. e(t)
u(t) z(t) + y(t)
PROCES +

Sl.1.
6. Perform the procedure as in 1, but the measurement error e (t) is a pseudorandom signal with a zero
average value and takes values in the range of -0.02 to 0.02.
7. Signals u(t), y(t) i e(t) for case from 5. Show graphicly

8. Using RLS and data generated 5 estimate values of process model.


e) RLS start on two different ways.
a1) RLS starts without any informations about parametars
a2) RLS starts in t=10 [s], depending on LS estimation until t=10 [s].
f) Graphicly show estipation of parametars in time 0 t 100 s .
g) Graphicly show error between real process and estimated values on output of process in time
0 t 100 s .
h) Compare results.

9. Compare results from 3 and 7.

You might also like