Stiff ODE - Numerical Analysis

import numpy as np
from matplotlib import pyplot as plt

Example 1

Consider the stiff ODE

y' = -100(y - \sin(t)), \qquad t \ge 0

(1)

y(0) = 1

(2)

The exact solution is

y(t) = \frac{10101}{10001}e^{-100t} - \frac{100}{10001}(\cos t - 100 \sin t)

(3)

def f(t,y):
    return -100.0*(y - np.sin(t))

def yexact(t):
    return 10101.0*np.exp(-100*t)/10001.0 - 100.0*(np.cos(t) - 100.0*np.sin(t))/10001.0

Forward Euler

y_n = y_{n-1} - 100 h [ y_{n-1} - \sin(t_{n-1})]

(4)

def ForwardEuler(t0,y0,T,h):
    N = int((T-t0)/h)
    y = np.zeros(N)
    t = np.zeros(N)
    t[0] = t0
    y[0] = y0
    for n in range(1,N):
        y[n] = y[n-1] + h*f(t[n-1],y[n-1])
        t[n] = t[n-1] + h
    return t,y

t0 = 0
y0 = 1
T  = 2*np.pi

h  = 0.95*(2.0/100.0)
t,y = ForwardEuler(t0,y0,T,h)

plt.plot(t,y,t,yexact(t))
plt.title("Forward Euler, h ="+str(h))
plt.legend(("Euler","Exact"))

h  = 1.0/100.0
t,y = ForwardEuler(t0,y0,T,h)

plt.plot(t,y,t,yexact(t))
plt.title("Forward Euler, h ="+str(h))
plt.legend(("Euler","Exact"));

Backward Euler scheme

y_n = y_{n-1} + h [ -100(y_n - \sin(t_n) ]

(5)

y_n = \frac{ y_{n-1} + 100 h \sin(t_n)}{1 + 100 h}

(6)

def BackwardEuler(t0,y0,T,h):
    N = int((T-t0)/h)
    y = np.zeros(N)
    t = np.zeros(N)
    t[0] = t0
    y[0] = y0
    for n in range(1,N):
        t[n] = t[n-1] + h
        y[n] = (y[n-1] + 100.0*h*np.sin(t[n]))/(1.0 + 100.0*h)
    return t,y

Trapezoidal scheme

y_n = y_{n-1} + \frac{h}{2}[ -100(y_{n-1}-\sin(t_{n-1})) - 100 (y_n - \sin(t_n))]

(7)

y_n = \frac{(1-50h)y_{n-1} + 50h(\sin(t_{n-1}) + \sin(t_n))}{1 + 50 h}

(8)

def Trapezoidal(t0,y0,T,h):
    N = int((T-t0)/h)
    y = np.zeros(N)
    t = np.zeros(N)
    t[0] = t0
    y[0] = y0
    for n in range(1,N):
        t[n] = t[n-1] + h
        y[n] = ((1.0-50.0*h)*y[n-1] + 50.0*h*(np.sin(t[n-1])+np.sin(t[n])))/(1.0 + 50.0*h)
    return t,y

We now run both backward Euler and trapezoidal scheme.

h  = 0.1
t,y = BackwardEuler(t0,y0,T,h)
plt.plot(t,y)

t,y = Trapezoidal(t0,y0,T,h)
plt.plot(t,y)

plt.plot(t,yexact(t))
plt.title("Step size h ="+str(h))
plt.legend(("Backward Euler","Trapezoid","Exact"));

Numerical Analysis

Asymptotic stability: Adams, BDF

Numerical Analysis

Stiff ODE: Forward Euler with variable step