Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 1

Estimation of Signals and Systems

Computer-based Assignment

Problem Statement:
Consider a single input single output system with the following transfer function:

5( + 2)( + 10)
() =
( + 1)( + 5)( + 20)

1. Determine the unit-step response of this system and plot the result for 100 samples
with a sampling time of 0.05 sec.
2. Discretize the above system with a sampling times of 0.05 sec, 0.1 sec and 0.2 sec
using the zero-order hold approximation and write down the discrete-time transfer
function in each case.
3. Using the data obtained in step 1, estimate the parameters of the discrete-time
model corresponding to the three sampling times given in step 2 and compare the
results with the true parameters obtained there. Use the least squares parameter
estimation technique employing (a) the matrix pseudo-inverse method and (b)
recursive parameter estimation technique.
4. Generate a sequence of 100 random numbers with zero mean and unit variance
obeying a Gaussian distribution. Plot the sequence.
5. Add the above random sequence with a multiplying factor K to the output
measurements obtained in step 1. Choose K =0.01, 0.05, 0.1, 0.2 and 0.5. Using this
noisy data, estimate the parameters of the discrete-time system corresponding to a
sampling time of 0.05 sec. Tabulate the results for various values of K and comment
on any bias in the parameter estimates.
6. Use the instrumental variables technique to overcome the bias in the parameter
estimates (if any) in step 5.
7. Use the generalized least squares technique to do the same.
8. Generate a pseudo-random binary sequence of length 100 and use it to excite the
system. Rework step 3 with this set of input-output data. Do you observe any
difference in the quality of the estimates?

You might also like