~

Behind the Laplace transform A

Generatlons of electdcal states that if x(t)is the input to a linear sys- tablished the missing rigor by showing how
enolneefs have valued thls tem with impulse response h(t), then the to interpret Heaviside’s work in terms of
rransfom as a fast mute output, y(t), is x(t) convolved with h(t) complex, Laplace-like integrals; but his sui-
to tlndlns the rrans/ent [Equation3, blackboard]. But if Y(s)andX(s) cide in 1929, before he could promulgate
as well as the steady. are the associated Laplace transforms, this these new ideas, probably accounts for his
stare outputs of clrcults imposing integral is completely tamed anonymity among EEs today.
havlno apertodlc as well [Equation 4, blackboard]. Until the end of the 1930s, the more ad-
as persodic Inputs In one of those astonishing coincidences vanced EEs continued to use Heaviside’s ad
that could mean mathematics is not just a hoc techniques. But lesser analysts were all
The Laplace transform is perhaps the mathe- game of made-up rules, the convolution in- too often swallowed up in the dangers of
matical signature of the electrical engineer, tegral also plays a central role in the theory methods a master like Heaviside could in-
having a long history of application to prob- of random variables: if x(t) and h(t) are the tuitively sense and sidestep.
lems of electrical engineering. It changes probability density functions (pdfs) of two Then, in 1937, the German mathematician
some of the most important differential independent random variables, X and H , Gustav Doetsch published his book Theorie
equations of physics into algebraic equations, then the pdf, y(t), of Y = X + H is found und Anwendung der h p h e - T r a m f m h i m .
which are generally far easier to solve. by convolving the pdfs of X and H . The That same year, L.A. Pipes published the
Despite its Gallic name, the transform Laplace transform consequently sees a lot first explicit application of the method to
originates with the Swiss mathematician, of use in probability theory, where Laplace electrical engineering problems in
Leonhard Euler (1707-1783), who in 1744 first found it. Philosophical Magazine, in a paper titled
wrote integrals that look much like the mod- Over a century had to elapse, though, be- “Laplacian Transform Circuit Analysis.”
ern version. These were adapted by the fore these sophisticated applications of the These works reduced Heaviside’s tricks to
Italian-French mathematical physicistJoseph Laplace transform could evolve from the routine methods. The technique quickly
Louis Lagrange (1736-1813) to the needs of form in which Laplace left it in 1827. For spread, and in the United States, Transients
probability theory, and his work in turn in- in Linear Systems: studied by the Laplace
fluenced the Frenchman Pierre Simon Transformation-the still classic text, known
Laplace (1749-1827). to generations of EEs as “Gardner and
By 1785 Laplace was writing the almost Barnes”-was published in 1942.
modem Laplace transform [Equation 1, More recently still, the development of
blackboard]. Today it is still used, as the computer programs that simulate even high-
Mellin transform, to solve certain differen- ly nonlinear circuits has threatened the prac-
tial equations with variable coefficients. But tical importance of the Laplace transform.
an electrical engineer would write it differ- Small enough to run on desktop PCs, yet
ently [Equation 2, blackboard]. powerful enough to provide the steady-state
Then in 1807 Laplace’s fellow- and transient behaviors of multistage elec-
countryman, Joseph Fourier, published the tronic circuits, such software packages have
first monograph describing the heat diffusion made the Laplace transform more important
equation. Intrigued, Laplace tried his hand for the theoretician than for the practical
at solving it, obtaining results that in turn in- designer and analyst. Will the Laplace trans-
spired Fourier’s discovery of his own EEs, the next big advance was made by the form (like stress analysis in bridge struts)be-
transform. eccentric Englishman Oliver Heaviside come one of those concepts learned in en-
The connection between the Fourier (1850-1925). He reduced differential equa- gineeering school but never used on the job?
transform (used by EEs to study the power tions directly to algebra by representing time TO PROBE FURTHER. A good guide to the early
and energy spectra of signals) and the differentiation as an operator. He usedp, as literature on the Laplace transform and
Laplace transform is intimate, but they are inpx = dx/dt andpZx = d2x/dt2,while 1Ip (x) Heaviside’s methods is Thomas J. Higgins,
not equivalent. For example, the step func- = the integral of x dt, and then manipulated “History of the Operational Calculus as
tion has a Laplace transform, but not a Fou- these equations using any algebraic trick he Used in Electric Circuit Analysis,” Electri-
rier transform. And while the Fourier trans- could think of, including his famous Heavi- cal Engineering, Vol. 68, January 1949,
form is useful in finding the steady-state side expansion theorem, which is essentially pp. 42-45.
output of a linear circuit in response to a the partial fraction expansion of modem Also see “Strange Mathematics,” Chap-
periodic input, the Laplace transform can Laplace theory. ter 10 of Oliver Heaviside:Sage In Solitude
provide both the steady-state and transient But Heaviside was notoriously uncon- by Paul J. Nahin (IEEE Press, New York,
responses for periodic and aperiodic inputs. cerned with rigor. For instance, he never 1988) and “Euler’s Version of the Laplace
For the modem EE, the lure of the blinked an eye when fractional operators, Transform” by Michael A.B. Deak in in The
Laplace transform is its ability to map the such asp”, arose. This lofty attitude of his American Mathematical Monthly, Vol. 87,
complicated operation of convolution into greatly delayed acceptance of his work, even April 1980, pp. 264-269.
multiplication. This integral has for decades though fractional operators were already old ABOUT THE AUTHOR. Paul J. Nahin is associate
driven electrical engineering undergraduates news in mathematics. professor of electrical engineering at the
to contemplate theology either for salvation Before he died, Heaviside exchanged University of New Hampshire in Durham.
or as an alternative career. The equation several letters with mathematician Thomas A member of the IEEE History Committee,
John I’Anson Bromwich (1875-1929) of Cam- his newest book is Time Machines: Time
Paul J. Nahin University of New Hampshire bridge University in England. Bromwich es- Travel in Physics and Science Fiction. +
60 0018-923519110003-0060$1.0001991IEEE IEEE SPECTRUM MARCH 1991

Authorized licensed use limited to: National Technical University of Athens. Downloaded on April 12,2010 at 15:47:58 UTC from IEEE Xplore. Restrictions apply.