LECTURE 8: AUTONOMOUS EQUATIONS MINGFENG ZHAO September 19, 2014 To solve differential equation M (x, y) + N (x, y) y 0 = 0: M y − Nx N 2) Compute the integrating factor µ(x): 1) Compute My (x, y), Nx (x, y) and µ(x) = e My −Nx N R dx . 3) Understand the meaning of the integrating factor, that is, [µM ] + [µN ] y 0 = 0 is exact, then there exists some φ(x, y) such that φx (x, y) = µ(x)M (x, y), and φy (x, y) = µ(x)N (x, y). 4) Since φx (x, y) = µ(x)M (x, y), then Z φ(x, y) = µ(x)M (x, y) dx + f (y). 5) Since φy (x, y) = µ(x)N (x, y), take the partial derivative with respect to y on the both sides of φ(x, y) = Z µ(x)M (x, y) dx + f (y), then Z µ(x)N (x, y) = φy (x, y) = µ(x)My (x, y) dx + f 0 (y). 6) Compute f (y), that is, Z f (y) = Z µ(x)N (x, y) − µ(x)My (x, y) dx dy. 7) Write down the solution: Z Z Z φ(x, y) = My (x, y) dx + µ(x)N (x, y) − µ(x)My (x, y) dx dy = C. Example 1. Solve ydx + (x2 y − x)dy = 0, x > 0. Rewrite the equation: y + (x2 y − x)y 0 = 0. 1 2 MINGFENG ZHAO Let M (x, y) = y and N (x, y) = x2 y − x, then and Nx = 2xy − 1. My = 1, So 2 − 2xy 2 M y − Nx = 2 =− . N x y−x x So the integrating factor is: µ(x) = e That is, the equation My −Nx N dx = e− R 2 x dx = e−2 ln x = 1 . x2 y x2 y − x 0 + y = 0 is exact. So there exist some φ(x, y) such that x2 x2 φx = Since φx = R y , x2 and φy = x2 y − x 1 =y− . 2 x x y , then x2 Z y y dx + f (y) = − + f (y). x2 x φ(x, y) = Take the partial derivative with respect to y, since φy = y − y− So f 0 (y) = y. We can take f (y) = 1 , then x 1 1 = φy = − + f 0 (y). x x y2 . So the solution is: 2 − y y2 + =C . x 2 Use Euler’s Method to approximate solutions of y 0 = f (x, y): Given the initial condition y(x0 ) = y0 and the step size h, compute the point (xk+1 , yk+1 ) from the previous point (xk , yk ) as follows: 1) Use the differential equation to compute the slope f (xk , yk ). 2) Calculate the next point (xk+1 , yk+1 ) using the formula xk+1 = xk + h yk+1 = yk + hf (xk , yk ) LECTURE 8: AUTONOMOUS EQUATIONS 3 Let xk = x0 + kh, and x1 = x0 + h y1 = y0 + hf (x0 , y0 ) x2 = x1 + h y2 = y1 + hf (x1 , y1 ) x3 = x2 + h .. . y3 = y2 + hf (x2 , y2 ) .. . xk+1 = xk + h .. . yk+1 = yk + hf (xk , yk ) .. . Theorem 1 (Picard’s Theorem on Existence and Uniqueness). If f (x, y) is continuous with respect to x and y, and ∂f (x, y) exists and is continuous near some (x0 , y0 ), then a solution to ∂y y 0 = f (x, y), y(x0 ) = y0 , exists for some small interval containing x0 , and is unique. AUTONOMOUS EQUATIONS An autonomous equation is of the form: dy = f (y). dx Then 1) y(x) ≡ a for some constant a such that f (a) = 0. In this case, y(x) ≡ a is called an equilibrium solution, and a is called a critical point of f . 2) If f (y) 6= 0, then Z x= 1 dy. f (y) For the slope field of an autonomous equation, the slopes are the same as long as they have same y-value. That is to say, if we know the slope field along a single vertical line(such single vertical line is called phase line/diagram/portrait), then we know the slope field in the entire xy-plane. On this vertical line, we mark all critical points, and then we draw arrows between critical points: If f (y) > 0, we draw an up arrow “↑”; If f (y) < 0, we draw a down arrow “↓”. Lemma 1. Let f be differentiable, f 0 is continuous, and y be a solution to y 0 = f (y). If y 0 (x0 ) = 0 for some x0 , then y(x) ≡ y(x0 ), for all x. 4 MINGFENG ZHAO Proof. Since y 0 (x0 ) = 0 and y 0 (x) = f (y(x)), then 0 = f (y(x0 )), which implies that y(x) ≡ y(x0 ) is a solution to y 0 = f (y). Since f is differentiable and f 0 is continuous, by Picard’s Theorem on Existence and Uniqueness, Theorem 1, we know that y(x) ≡ y(x0 ), for all x. Theorem 2. Let f be differentiable, f 0 is continuous, and y be a solution to y 0 = f (y), then either y 0 (x) ≡ 0 for all x, or, y 0 (x) > 0 for all x, or y 0 (x) < 0 for all x. Proof. If y 0 (x) ≡ 0 for all x, we are done. If y 0 (x) 6≡ 0, by Lemma 1, we must have that either y 0 (x) > 0 for all x, or y 0 (x) < 0 for all x. Remark 1. Theorem 2 tells us that any solution to an autonomous equation must be either constant function, or strictly increasing, or strictly decreasing. Example 2. Find equilibrium points of y 0 = y(1 − y), draw the phase diagram, and describe the solutions y1 , y2 , y3 , y4 satisfy the following initial conditions: 1) y1 (0) = 0, 2) y(0) = −1, 3) y(0) = 2, 4) y(0) = 1 . 2 Let f (y) = y(1 − y). Solve f (y) = y(1 − y) = 0, then y = 0 or y = 1, that is, all equilibrium points of y 0 = y(1 − y) are: 0 and 1. For the initial value problems: 1) Since f (0) = 0(1 − 0) = 0, then y1 is a constant function, that is, y1 (x) ≡ 0. 2) Since f (−1) = (−1)[1 − (−1)] = −2 < 0, then y2 (x) is strictly decreasing . 3) Since f (2) = 2(1 − 2) = −2 < 0, then y3 (x) is strictly decreasing. LECTURE 8: AUTONOMOUS EQUATIONS 1 1 1 1 4) Since f ( ) = (1 − ) = > 0, then 2 2 2 4 y4 (x) is strictly increasing. Figure 1. Phase diagram of y 0 = y(1 − y) Figure 2. Slop field of y 0 = y(1 − y) 5 6 MINGFENG ZHAO Figure 3. Slop field of y 0 = y(1 − y) Remark 2. In general, for any initial day y(x0 ) = y0 if x0 is between two critical points, then the solution must be the whole real line, like the case y(0) = 1/2 in Example 2. If x0 is not between two critical points, for example, consider the problem: y 0 = y 2 , y(0) = 1. It’s easy to see that 1 . So the solution interval is (−∞, 1). However, we say that y(x) → ∞ as x → 1− . the solution is: y(x) = 1−x Definition 1. Consider the autonomous equation: dy = f (y). dx Let a be a critical point of f , that is, f (a). Then I. We say a is stable or a sink if any solution with initial condition close to a is asymptotic to a as x increases. Figure 4. Stable or sink LECTURE 8: AUTONOMOUS EQUATIONS 7 II. We say a is a source if all solutions start close to a tend toward y as x decreases, and tend away from a as x increases. Figure 5. Source III. We say a is a node if the equilibrium point a is neither a source nor a sink. Figure 6. Node If a is either a source or a node, we say a is unstable. Example 3. Consider Logistic population mode: dP = kP dt P 1− N , where k, N > 0 are constants. 1) Equilibrium points: P = 0 and P = N . 2) Determine the signs: P > N, 0 < P < N, P < 0, dP = (+)(−) < 0 dt dP = (+)(+) > 0 dt dP = (−)(+) < 0 dt ↓ ↑ ↓ 3) Types of equilibrium points: 0 is a source (unstable); N is a sink (stable) . 8 MINGFENG ZHAO 4) Initial value problems: – If the initial value is P (0) = N , then P (t) is strictly increasing, P (t) → N as t → ∞, and P (t) → 0 as 3 t → −∞. – If the initial value is P (0) = N , then P (t) ≡ N for all t, i.e., the population will stay as N . – If the initial value is P (0) = N + 5, then P (t) is strictly decreasing, P (t) → N as t → ∞, and P (t) → ∞ as t → −∞. Theorem 3 (Linearization Theorem). Suppose a is an equilibrium point of the differential equation: dy = f (y), dx where f is differentiable and f 0 is continuous. Then 1) If f 0 (a) < 0, then a is a sink (stable). 2) If f 0 (a) > 0, then a is a source (unstable). 3) If f 0 (a) = 0, then undecidable. (Need more information to decide the type of a) Example 4. y 0 = (2 − y) sin(y). Let f (y) = (2 − y) sin(y), then 1) Equilibrium points: y = 2, y = 0, ±π, ±2π, · · · . 2) Types of equilibrium points: Since f 0 (y) = − sin(y) + (2 − y) cos(y), then f 0 (2) = − sin(2) < 0, f 0 (0) = 2 > 0 For any positive integer k, we have f 0 (2kπ) f 0 (−2kπ) f 0 ((2k − 1)π) = (2 − 2kπ) cos(2kπ) = 2 − 2kπ < 0 = (2 + 2kπ cos(−2kπ) = 2 + 2k > 0 = [2 − (2k − 1)π] cos((2k − 1)π) = −[2 − (2k − 1)π] LECTURE 8: AUTONOMOUS EQUATIONS f 0 (−(2k − 1)π) = (2k − 1)π − 2 > 0 = [2 + (2k − 1)π] cos(−(2k − 1)π) 9 = −[2 + (2k − 1)π] = −2 − (2k − 1)π < 0. Then P = 2, 2kπ, −(2k − 1)π are sink (stable), P = 0, −2kπ, (2k − 1)π are source (unstable), k is any positive integer. . 3) Initial value problems: – If y(0) = 0.4 ∈ (0, 2), then y(x) is strictly increasing, y(x) → 2 as x → ∞, and y(x) → 0 as x → −∞. – If y(0) = 3 ∈ (2, π), then y(x) is strictly decreasing, y(x) → 2 as x → ∞, and y(x) → π as x → −∞. Department of Mathematics, The University of British Columbia, Room 121, 1984 Mathematics Road, Vancouver, B.C. Canada V6T 1Z2 E-mail address: mingfeng@math.ubc.ca