Download presentation
Presentation is loading. Please wait.
Published byNathalie Molenaar Modified over 5 years ago
1
DIGITAL CONTROL SYSTEM WEEK 3 NUMERICAL APPROXIMATION
Given the set of discrete values π ππ , π πβ1 π ,β¦,π 0 , where π= 1,2,3,.., one defines the ππππ π‘βπππππ€πππ Difference π»π ππ =π ππ βπ πβ1 π This permits its solution on a digital computer. Dr. Kalyana Veluvolu
2
π π+1 π βπ ππ = π π+1 π +π(ππ) 2 π
Integral approximation of π π‘ = 0 π‘ π π ππ First forward difference (Eulerβs Technique) π π+1 π βπ ππ =ππ ππ Trapezoidal: π π+1 π βπ ππ = π π+1 π +π(ππ) 2 π Dr. Kalyana Veluvolu
3
Example: Evaluate π π‘ = 0 π‘ π π ππ by a numerical integration technique π ππ = 0 ππ π π ππ β π=0 πβ1 π ππ βπ‘= π=0 πβ1 ππ(ππ) Dr. Kalyana Veluvolu
4
Notice that due to finite word length,
Decreasing T increases N and hence the computation time required and increases the roundoff and truncation errors by use of recursive equation. If it is not small enough, aliasing or frequency folding problems will exist. Thus the selection of T is based on trade-off when all factors are considered Dr. Kalyana Veluvolu
5
SAMPLING AND SIGNAL RECONSTRUCTION (for Ideal Samplers)
Continuous-time signals often need to be sampled for storage, transmission and processing. Applications are found in control and communications. Here we study how to sample a continuousβtime signal, how to analyze a sampled signal, and how to reconstruct a continuous-time signal from its sampled version. Dr. Kalyana Veluvolu
6
Where T is the sampling period
Sampling: Continuous-time signal: x(t), -β<π‘<β Sampled signal: x(nT), n = 0 ,Β±1,Β±2, β¦ Where T is the sampling period Dr. Kalyana Veluvolu
7
Representation of sampled signal using impulse train function: Impulse train function: π π‘ = π=ββ β πΏ(π‘βππ) Sampled signal: π₯ π π‘ =π₯ π‘ π π‘ = π=ββ β π₯ π‘ πΏ π‘βππ = π=ββ β π₯(ππ)πΏ(π‘βππ) Dr. Kalyana Veluvolu
8
Fourier transform of sampled signal: Fourier series of pulse train function: Since p(t) is periodic, it can be expressed as Fourier series as below; π π‘ = π=ββ β π π π ππ π π π‘ Where π π = 2π π is the sampling frequency in rad/sec. and π π = 1 π βπ 2 π 2 π(π‘) π βπ π π π‘ ππ‘= 1 π βπ 2 π 2 πΏ π‘ π βππ π π ππ‘= 1 π Dr. Kalyana Veluvolu
9
And the sampled signal is rewritten as
Hence, π π‘ = π=ββ β 1 π π ππ π π π‘ And the sampled signal is rewritten as π₯ π π‘ =π₯ π‘ π π‘ = π=ββ β 1 π π₯(π‘) π ππ π π π‘ Using the property of multiplication by complex exp.: π π π = π=ββ β 1 π π(πβπ π π ) Sum of shifted Fourier transforms Dr. Kalyana Veluvolu
10
Graphical illustration:
Dr. Kalyana Veluvolu
11
Aliasing: From the figure in previous page, we see that the shifted copies of π(π) do not overlap if the sampling frequency is at least twice of the signal bandwidth, i.e.: π π >2π΅ If the sampling frequency is too low or the signal bandwidth is too high, it will result in aliasing (overlap of the shifted copies of π(π)) as below. Dr. Kalyana Veluvolu
12
Signal Reconstruction:
From the analysis above, we know that Fourier transform of the original signal, π(π), can be perfectly reconstructed from the spectrum of the sampled signal if and only if To do so, we simply need to apply an ideal lowpass filter to cut the copies of π(π) in π π (π). Since π(π) and x(t) have a one-to-one relationship, we can also recover x(t) if and only if the above criterion is satisfied. This criterion is the famous Nyquist sampling criterion or Shannon sampling theorem. π π >ππ© Dr. Kalyana Veluvolu
13
Example: Consider the continuous-time signal
π(π‘)=sinβ‘( π 1 π‘+π) Which is sampled with a sampling time of π 1 = π π , Where 0< π 1 < π π 2 .Thus letting π‘=ππ yields π π‘ = sin π 1 ππ+π Next consider the continuous-time signal π π‘ = sin π 1 + 2ππ π π‘+π Where j = 1,2,3... And which is sampled with the same sampling time T. π ππ = sin π 1 + 2ππ π ππ+π = sin ππ π 1 +2πππ+π =sinβ‘(ππ π 1 +π) Dr. Kalyana Veluvolu
14
Comparing the equations reveals that they are identical
Comparing the equations reveals that they are identical. Thus, it is impossible to differentiate between two sampled sinusoids whose radian frequencies differ by an integral multiple of 2π π . It must be avoided by selecting small-enough T in accordance with the Shannon sampling theorem π π >2 π π ππ π π > π π Effectively, all radian frequencies are then folded into the interval 0< π 1 < π π 2 .The frequency 2π π rad/s is commonly referred to as the folding Dr. Kalyana Veluvolu
15
Example: Voice signals over telephone lines
Analog voice signals need to be sampled for digital transmission. A bandwidth of 4kHz is allocated for digital transmission. The analog voice signals has signal bandwidth up to 20kHz or so. So it needs to be lowpass filtered before sampling in order to avoid aliasing (i.e. overlapping of high frequencies signals with low frequency signals). Determine the cutoff frequency of the lowpass filter and the sampling frequency. Answer: cutoff freq <=4KHz, Sampling freq = 8kHz Dr. Kalyana Veluvolu
16
Formula for signal reconstruction (assuming no aliasing):
Proof: we may recover π(π) using the following lowpass filter. π» π = π, 0, Its impulse response is given by (see F.T. pairs table) β π‘ = π΅π π π πππ π΅ π π‘ π₯ π‘ = π΅π π π=ββ β π₯ ππ π πππ π΅ π (π‘βππ) βπ΅β€πβ€B otherwise Dr. Kalyana Veluvolu
17
The F.T of the filtered output is given by π(π)=π»(π) π π (π) Using the property of convolution, we get π¦ π‘ =β π‘ β π₯ π π‘ = ββ β π₯ π π β π‘βπ ππ Recall: π₯ π π‘ =π₯ π‘ π π‘ = π=ββ β π₯(ππ)πΏ(π‘βππ) We have : π¦ π‘ = ββ β ββ β π₯ ππ πΏ πβππ β π‘βπ ππ ββ β π₯(ππ) ββ β πΏ πβππ β π‘βπ ππ = ββ β π₯ ππ β π‘βππ = π΅π π π=ββ β π₯ ππ π πππ π΅ π (π‘βππ) = Dr. Kalyana Veluvolu
18
NON-IDEAL SAMPLERS π π β π‘ =π π‘ π π‘ A modulation process:
Dr. Kalyana Veluvolu
19
Sampled signal: Dr. Kalyana Veluvolu
20
Dr. Kalyana Veluvolu
21
IDEAL SAMPLER & LAPLACE TRANSFORM
If the duration πΎ of the sampling pulse is much less than the sampling time T and the smallest time constant of e(t), the output of the pulse modulator can be approximated as an ideal impulse train. π π β π‘ β 1 πΎ π=0 β π ππ [π’ π‘βππ βπ’(π‘βππβπΎ)] Where π π β ππ = π(ππ) π πΎ 0 πππ ππβ€π‘β€ππ+ π πΎ πππ ππ+πβ€π‘β€ππ+1 k = 0,1,2,3,β¦. Dr. Kalyana Veluvolu
22
Taking the Laplace transformation, πΈ π β(π )β π=0 β π ππ 1β π βπΎπ πΎπ π βπππ Assuming πΎ is very small, such that 1β π βπΎπ = 1β[1βπΎπ + πΎπ 2 2! β πΎπ 3 3! +β¦]βπΎπ Then πΈ β π = π=0 β π(ππ) π βπππ Dr. Kalyana Veluvolu
23
Z TRANSFORM The Laplace transform of a sampled function e*(t) is πΈ β π = π=0 β π(ππ) π βπππ Let π§β‘ π ππ π = 1 π lnz We have the z transform: πΈ π§ = πΈ β (π ) π§= π ππ = π=0 β π(ππ) π§ βπ That is, πΈ π§ =π π β π‘ = πΈ β (π ) π§= π ππ Dr. Kalyana Veluvolu
24
πΈ π§ = πΎ π§β π§ 1 π§β π§ 2 β¦(π§β π§ π€ ) π§β π 1 π§β π 2 β¦(π§β π π )
The z-transform as defined above is also called one-sided z-transform because the summation starts from 0. For most practical application, the one sided z-transform has a closed-form representation in its region of convergence. πΈ π§ = πΎ( π§ π€ + π π€β1 π§ π€β1 +β¦+ π 1 π§+ π 0 ) π§ π + π πβ1 π§ πβ1 +β¦+ π 1 π§+ π 0 = πΎ( π§ βπ+π€ +β¦+ π 1 π§ βπ+1 + π 0 π§ βπ ) 1+ π πβ1 π§ β1 +β¦+ π 1 π βπ+1 + π 0 π§ βπ Factoring the polynomials, yields πΈ π§ = πΎ π§β π§ 1 π§β π§ 2 β¦(π§β π§ π€ ) π§β π 1 π§β π 2 β¦(π§β π π ) Where π π and π§ π are the poles and zeros of E(z) Dr. Kalyana Veluvolu
25
Example: Given π π‘ =π’ π‘ , determine πΈ(π§)
π β π‘ = π=0 β πΏ π‘βππ = πΏ π‘ + πΏ π‘βπ + πΏ π‘β2π +β¦ πΈ β π =πΏ π β π‘ = π=0 β π βπππ =1+ π βππ + π β2ππ + β¦ The z-transform is given by πΈ π§ = π=0 β π§ βπ =1+ π§ β1 + π§ β2 + π§ β3 +β¦ It is an open-form expression of E(z). The closed form is πΈ π§ = 1 1β π§ β1 = π§ π§β1 Note that the region of convergence is π§ >1. Dr. Kalyana Veluvolu
26
Dr. Kalyana Veluvolu
27
Converting Laplace transform to z-transform Step 1: Perform a partial-fraction expansion on E(s). πΈ π = π΄ 0 (π + πΌ 0 ) + π΄ 1 (π + πΌ 1 ) +β¦+ π΄ π (π + πΌ π ) Step 2: Convert each term using the Transforms Table. πΈ π§ =π πΈ π =π π΄ 0 (π + πΌ 0 ) +π π΄ 1 (π + πΌ 1 ) +β¦+π π΄ π (π + πΌ π ) Dr. Kalyana Veluvolu
28
Example: Determine E(z) for the following:
πΈ π = 5 π (π +1)(π +5) Step 1: Partial fraction expansion πΈ π = 1 π β π (π +5) Step 2: Convert each term πΈ π§ = π§ π§β1 β 1.25π§ π§β π βπ π§ (π§β π β5π ) Dr. Kalyana Veluvolu
29
SUMMARY Digital representation of continuous-time signals
Linear time-invariant systems Sampled-data systems Sampling and signal reconstruction Shannon (Nyquist) sampling criterion Aliasing Laplace transform of sampled signals Z-transform Dr. Kalyana Veluvolu
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.