The comparison principle allows us to bound solutions to differential equations without explicitly solving them. It states that if the derivative of a function v(t) is bounded above by the derivative of the solution u(t) to a related differential equation, then v(t) will always be less than or equal to u(t). This lemma is proved by introducing an auxiliary equation and showing v(t) is bounded by its solution, and then bounded by u(t).
The comparison principle allows us to bound solutions to differential equations without explicitly solving them. It states that if the derivative of a function v(t) is bounded above by the derivative of the solution u(t) to a related differential equation, then v(t) will always be less than or equal to u(t). This lemma is proved by introducing an auxiliary equation and showing v(t) is bounded by its solution, and then bounded by u(t).
The comparison principle allows us to bound solutions to differential equations without explicitly solving them. It states that if the derivative of a function v(t) is bounded above by the derivative of the solution u(t) to a related differential equation, then v(t) will always be less than or equal to u(t). This lemma is proved by introducing an auxiliary equation and showing v(t) is bounded by its solution, and then bounded by u(t).
The comparison principle allows us to bound solutions to differential equations without explicitly solving them. It states that if the derivative of a function v(t) is bounded above by the derivative of the solution u(t) to a related differential equation, then v(t) will always be less than or equal to u(t). This lemma is proved by introducing an auxiliary equation and showing v(t) is bounded by its solution, and then bounded by u(t).
Copyright:
Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online from Scribd
Download as pdf or txt
You are on page 1/ 3
Comparison Principle
Quite open when we study the state equation x = f (t, x) we need to com pute bounds on the solution x(t) without computing the solution itself. The Gronwall- Bellman inequality (Lemma 2.1) is one tool that can be used toward that goal. Another tool is the comparison lemma. It applies to a situation where the derivative of a scalar dierentiable function v(t) satises inequality of the form v f (t, v(t)) for all t in a certain time interval. Such inequality is called a dierential inequality and a function v(t) satisfying the inequality is called a solution of the dierential inequality. The comparison lemma compares the solution of the dierential inequality v(t) f (t, v(t)) with the solution of the dierential equation u = f (t, u). The lemma applies even when v(t) is not dierentiable, but has an upper right-hand derivative D+ v(t), which satises a dierential inequality. The upper right-hand derivative D+ v(t) is dened in Appendix C.2. For our purpose, it is enough to know two facts: If v(t) is dierentiable at t, then D+ v(t) = v(t). If 1 |v(t + h) v(t)| g(t, h), h and h0+
h (o, b]
lim g(t, h) = g0 (t)
then D+ v(t) g0 (t). The limit h 0+ means that h approaches zero from above. Lemma 2.5 (Comparison Lemma) Consider the scalar dierential equation u = f (t, u), 1 u(t0 ) = u0
where f (t, u) is continuous in t and locally Lipschitz in u, for all t 0 and all u J R. Let [t0 , T ) (T could be innity) be the maximal interval of existence of the solution u(t), and suppose u(t) J for all t [t0 , T ). Let v(t) be a continuous function whose upper right-hand derivative D+ v(t) satises the dierential inequality D+ v(t) f (t, v(t)), v(t0 ) u0
with v(t) J for all t [t0 , T ). Then v(t) u(t) for all t [t0 , T ). Proof: See Appendix C.2
Appendix C.2: Proof of Lemma 2.5
The upper right-hand derivative D+ v(t) is dened by D+ v(t) = lim sup h0+
v(t + h) v(t) h
where lim supn (the limit superior) of a sequence of real numbers {xn } is a real number y satisfying the following two conditions: for every xn < y + ; given > 0 and m > 0, there exists an integer n > m such that > 0, there exists an integer N such that n > N implies
xn > y . The rst statement means that ultimately all terms of the sequence are less than y + , and the second one means that innitely many terms are greater than y . One of the properties of lim sup is that if zn xn for each n = 1, 2, ..., then lim supn zn lim supn xn . From this property, we see that if |v(t + h) v(t)|/h g(t, h), h (0, b] and limh0+ g(t, h) = g0 (t), then D+ v(t) g0 (t). To prove lemma 2.5, consider the dierential equation z = f (t, z) + , z(t0 ) = u0 2 (C.5)
where is a positive constant. On any compact interval [t0 , t1 ], we conclude that for any > 0, there is > 0 such that if < , then (C.5) has a unique solution z(t, ) dened on [t0 , t1 ] and |z(t, ) u(t)| < , t [t0 , t1 ] (C.6)
Claim 1: v(t) z(t, ) for all t [t0 , t1 ]. This claim can be shown by contradiction, for if it were not true, there would be times a, b (t0 , t1 ] such that v(a) = z(a, ) and v(t) > z(t, ) for a < t b. Consequently, v(t) v(a) > z(t, ) z(a, ), which implies D+ v(a) z(a, ) = f (a, z(a, )) + > f (a, v(a)) which contradicts the inequality D+ v(t) f (t, v(t)). Claim 2: v(t) u(t) for all t [t0 , t1 ]. Again, this claim can be shown by contradiction, for if it were not true, there would exist a (t0 , t1 ] such that v(a) > u(a). Taking and using (C.6), we obtain v(a) z(a, ) = v(a) u(a) + u(a) z(a, ) which contradits the statement of Claim 1. Thus, we have shown that v(t) u(t) for all t [t0 , t1 ]. Since this is true on every compact interval, we conclude that it holds for all t t0 . If it was not the case, let T < be the rst time the inequality is violated. We have v(t) u(t) for all t [t0 , T ) and, by continuity, v(T ) = u(T ). Consequently, we can extend the inequality to the interval [T, T + ] for some > 0, which contradicts the claim that T is the rst time the inequality is violated. 3 = [v(a) u(a)]/2 t (a, b]
