This preview shows page 1-2-3-4-5-6-7-8-9-10-72-73-74-75-76-77-78-79-80-81-82-146-147-148-149-150-151-152-153-154-155 out of 155 pages.
18.03 Supplementary Notes Spring 2010 c∞ Haynes R. Miller and MIT, 2004, 2006, 2008. 2010Contents 0. Preface 1 1. Notation and language 3 2. Modeling by first order linear ODEs 6 3. Solutions of first order linear ODEs 10 4. Sinusoidal solutions 16 5. The algebra of complex numbers 23 6. The complex exponential 27 7. Beats 34 8. RLC circuits 38 9. Normalization of solutions 41 10. Operators and the exponential response formula 45 11. Undetermined coefficients 53 12. Resonance and the exponential shift law 55 13. Natural frequency and damping ratio 60 14. Frequency response 62 15. The Wronskian 72 16. More on Fourier series 75 17. Impulses and generalized functions 86 18. Impulse and step responses 93 19. Convolution 98 20. Laplace transform technique: coverup 101 21. The Laplace transform and generalized functions 106 22. The pole diagram and the Laplace transform 112 23. Amplitude response and the pole diagram 119 24. The Laplace transform and more general systems 121 25. First order systems and second order equations 123 26. Phase portraits in two dimensions 1270. Preface This packet collects notes I have produced while teaching 18.03, Or-dinary Differential Equations, at MIT in 1996, 1999, 2002, 2004, 2006, 2008, and 2010. They are intended to serve several rather different purposes, supplementing but not replacing the course textbook. In part they try to increase the focus of the course on topics and perspectives which will be found useful by engineering students, while maintaining a level of abstraction, or breadth of perspective, sufficient to bring into play the added value that a mathematical treatment offers. For example, in this course we use complex numbers, and in partic-ular the complex exponential function, more intensively than Edwards and Penney do, and several of the sections discuss aspects of them. This ties in with the “Exponential Response Formula,” which seems to me to be a linchpin for the course. It leads directly to an understanding of amplitude and phase response curves. It has a beautiful extension covering the phenomenon of resonance. It links the elementary theory of linear differential equations with the use of Fourier series to study LTI system responses to periodic signals, and to the weight function appearing in Laplace transform techniques. It allows a direct path to the solution to standard sinusoidally driven LTI equations which are often solved by a form of undetermined coefficients, and leads to the ex-pression of the sinusoidal solution in terms of gain and phase lag, more useful and enlightening than the expression as a linear combination of sines and cosines. As a second example, I feel that the standard treatments of Laplace transform in ODE textbooks are wrong to sacrifice the conceptual con-tent of the transformed function, as captured by its pole diagram, and I discuss that topic. The relationship between the modulus of the trans-fer function and the amplitude response curve is the conceptual core of the course. Similarly, standard treatments of generalized functions, impulse response, and convolution, typically all occur entirely within the context of the Laplace transform, whereas I try to present them as useful additions to the student’s set of tools by which to represent natural events. In fact, a further purpose of these notes is to try to uproot some aspects of standard textbook treatments which I feel are downright misleading. All textbooks give an account of beats which is mathe-matically artificial and nonsensical from an engineering perspective. I give a derivation of the beat envelope in general, a simple and revealinguse of the complex exponential. Textbooks stress silly applications of the Wronskian, and I try to illustrate what its real utility is. Text-books typically make the theory of first order linear equations seem quite unrelated to the second order theory; I try to present the first or-der theory using standard linear methods. Textbooks generally give an inconsistent treatment of the lower limit of integration in the definition of the one-sided Laplace transform, and I try at least to be consistent. A final objective of these notes is to give introductions to a few top-ics which lie just beyond the limits of this course: damping ratio and logarithmic decrement; the L2 or root mean square distance in the the-ory of Fourier series; the exponential expression of Fourier series; the Gibbs phenomenon; the Wronskian; a discussion of the ZSR/ZIR de-composition; the Laplace transform approach to more general systems in mechanical engineering; and a treatment of a class of “generalized functions,” which, while artificially restrictive from a mathematical perspective, is sufficient for all engineering applications and which can be understood directly, without recourse to distributions. These essays are not formally part of the curriculum of the course, but they are written from the perspective developed in the course, and I hope that when students encounter them later on, as many will, they will think to look back to see how these topics appear from the 18.03 perspective. I want to thank my colleagues at MIT, especially the engineering fac-ulty, who patiently tutored me in the rudiments of engineering: Steve Hall, Neville Hogan, Jeff Lang, Kent Lundberg, David Trumper, and Karen Willcox, were always on call. Arthur Mattuck, Jean Lu, and Lindsay Howie read early versions of this manuscript and offered frank advice which I have tried to follow. I am particularly indebted to Arthur Mattuck, who established the basic syllabus of this course. He has patiently tried to tutor me in how to lecture and how to write (with only moderate success I am afraid). He also showed me the approach to the Gibbs phenomenon included here. My thinking about teach-ing ODEs has also been influenced by the the pedagogical wisdom and computer design expertise of Hu Hohn, who built the computer manipulatives (“Mathlets”) used in this course. They can be found at http://www-math.mit.edu/daimp. Assorted errors and infelicities were caught by students in 18.03 and by Professor Sridhar Chitta of MIST, Hyderabad, India, and I am grateful to them all. Finally, I am happy to record
View Full Document
Unlocking...