563489578934
694 Probability and Random Variables Appendix B However, you might ask, “What is the significance of the mean, variance, and other moments in electrical engineering problems?” In Chapter 6, it is shown that if x represents a voltage or current waveform, the mean gives the DC value of the waveform. The second moment (r = 2) taken about the origin (x 0 = 0), which is x 2 , gives the normalized power. s 2 gives the normalized power in the corresponding AC coupled waveform. Consequently, 3x 2 is the RMS value of the waveform, and s is the RMS value of the corresponding ac coupled waveform. In statistical terms, m gives the center of gravity of the PDF, and s gives us the spread of the PDF about this center of gravity. For example, in Fig. B–5 the voltage distribution for a collection of flashlight batteries is given. The mean is xq = 1.25 V, and the standard deviation is s = 0.25 V. This figure illustrates a Gaussian distribution that will be studied in detail in Sec. B–7. For this Gaussian distribution, the area under f(x) from x = 1.0 to 1.5 V, which corresponds to the interval xq ; s, is 0.68. Thus, we conclude that 68% of the batteries have voltages within one standard deviation of the mean value (Gaussian distribution). There are several ways to specify a number that is used to describe the typical, or most common, value of x. The mean m is one such measure that gives the center of gravity. Another measure is the median, which corresponds to the value x = a, where A third measure is called the mode, which corresponds to the value of x where f(x) is a maximum, assuming that the PDF has only one maximum. For the Gaussian distribution, all these measures give the same number, namely, x = m. For other types of distributions, the values obtained for the mean, median, and mode will usually be nearly the same number. The variance is also related to the second moment about the origin and the mean, as described by the following theorem: THEOREM. s 2 = x 2 - (xq) 2 F(a) = 1 2 . A proof of this theorem illustrates how the ensemble average operator notation is used. Proof. s 2 = (x - xq) 2 (B–29) = [x 2 ] - [2xxq] + [(xq)] 2 (B–30) Because [ # ] is a linear operator, [2xxq] = 2xqxq = 2(xq) 2 . Moreover, (xq) 2 is a constant, and the average value of a constant is the constant itself. That is, for the constant c, q q cq = cf(x) dx = c f(x) dx = c L L -q -q (B–31) So, using Eq. (B–31) in Eq. (B–30), we obtain which is equivalent to Eq. (B–29). s 2 = x 2 - 2(xq) 2 + (xq) 2
- Page 1386: A p p e n d i x MATHEMATICAL TECHNI
- Page 1390: A-4 Integral Calculus 671 d sin ax
- Page 1394: A-5 Integral Tables 673 L x2 cos x
- Page 1398: A-8 The Dirac Delta Function 675 A-
- Page 1402: A-9 Tabulation of Sa(x) = (sin x)/x
- Page 1406: A-10 Tabulation of Q (z) 679 z Q(z)
- Page 1410: B-2 Sets 681 used when conversing w
- Page 1414: B-3 Probability and Relative Freque
- Page 1418: B-5 Cumulative Distribution Functio
- Page 1422: B-5 Cumulative Distribution Functio
- Page 1426: B-5 Cumulative Distribution Functio
- Page 1430: B-5 Cumulative Distribution Functio
- Page 1434: B-6 Ensemble Average and Moments 69
- Page 1440: TABLE B-1 SOME DISTRIBUTIONS AND TH
- Page 1444: 698 Probability and Random Variable
- Page 1448: 700 Probability and Random Variable
- Page 1452: 702 Probability and Random Variable
- Page 1456: ƒ ƒ 704 Probability and Random Va
- Page 1460: 706 Probability and Random Variable
- Page 1464: 708 Probability and Random Variable
- Page 1468: 710 3. F(a 1 , a 2 ,..., a N ) Prob
- Page 1472: 712 Probability and Random Variable
- Page 1476: 714 Probability and Random Variable
- Page 1480: 716 Probability and Random Variable
- Page 1484: 718 Probability and Random Variable
694<br />
Probability and Random Variables<br />
Appendix B<br />
However, you might ask, “What is the significance of the mean, variance, and other moments<br />
in electrical engineering problems?” In Chapter 6, it is shown that if x represents a voltage or<br />
current waveform, the mean gives the DC value of the waveform. The second moment (r = 2)<br />
taken about the origin (x 0 = 0), which is x 2 , gives the normalized power. s 2 gives the normalized<br />
power in the corresponding AC coupled waveform. Consequently, 3x 2 is the RMS value of<br />
the waveform, and s is the RMS value of the corresponding ac coupled waveform.<br />
In statistical terms, m gives the center of gravity of the PDF, and s gives us the spread<br />
of the PDF about this center of gravity. For example, in Fig. B–5 the voltage distribution for a<br />
collection of flashlight batteries is given. The mean is xq = 1.25 V, and the standard deviation<br />
is s = 0.25 V. This figure illustrates a Gaussian distribution that will be studied in detail in<br />
Sec. B–7. For this Gaussian distribution, the area under f(x) from x = 1.0 to 1.5 V, which<br />
corresponds to the interval xq ; s, is 0.68. Thus, we conclude that 68% of the batteries have<br />
voltages within one standard deviation of the mean value (Gaussian distribution).<br />
There are several ways to specify a number that is used to describe the typical, or<br />
most common, value of x. The mean m is one such measure that gives the center of gravity.<br />
Another measure is the median, which corresponds to the value x = a, where<br />
A third measure is called the mode, which corresponds to the value of x where f(x) is a maximum,<br />
assuming that the PDF has only one maximum. For the Gaussian distribution, all<br />
these measures give the same number, namely, x = m. For other types of distributions, the<br />
values obtained for the mean, median, and mode will usually be nearly the same number.<br />
The variance is also related to the second moment about the origin and the mean, as<br />
described by the following theorem:<br />
THEOREM.<br />
s 2 = x 2 - (xq) 2 F(a) = 1 2 .<br />
A proof of this theorem illustrates how the ensemble average operator notation is used.<br />
Proof.<br />
s 2 = (x - xq) 2<br />
(B–29)<br />
= [x 2 ] - [2xxq] + [(xq)] 2<br />
(B–30)<br />
Because [ # ] is a linear operator, [2xxq] = 2xqxq = 2(xq) 2 . Moreover, (xq) 2 is a constant, and the<br />
average value of a constant is the constant itself. That is, for the constant c,<br />
q<br />
q<br />
cq = cf(x) dx = c f(x) dx = c<br />
L L<br />
-q<br />
-q<br />
(B–31)<br />
So, using Eq. (B–31) in Eq. (B–30), we obtain<br />
which is equivalent to Eq. (B–29).<br />
s 2 = x 2 - 2(xq) 2 + (xq) 2