MAT323(Dynamical Systems) Assignment 2 Ahnaf Shahriar Asif (22301787)

2.3.3

The growth of cancerous tumors can be modeled by the Gompertz law Ṅ = −aN ln (bN ), where N (t) is
proportional to the number of cells in the tumor, and a, b > 0 are parameters.

(a)

Let Ṅ = 0
−aN ln(bN ∗ ) = 0
∗

N ∗ = 0 or ln(bN ∗ ) = 0
1
bN ∗ = 1 =⇒ N ∗ =
b
So, the fixed points are N ∗ = 0 and N ∗ = 1b .

The below graph is plotted where y = Ṅ and a = 1.5, b = 3.

0.20
1/b
0.15

0.10

0.05

0.00

0.05

0.10

0.15

0.20
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

1
From the figure above, we can see that b is a stable fixed point and 0 is an unstable point. If we decrease
the value of a, the graph gets steeper.

1
So, a represents the growth frequency, and b represents the stable size of the tumor.

(b)

Ṅ = −aN ln (bN )
dN
= −aN ln (bN )
dt
dN
= −adt
N ln (bN )
dN
If we Let x = ln (bN ) which implies dx = N . . . (1)
Applying (1),

dx
= −adt
Z x Z
1
dx = −adt
x
ln|x| = −at + c
ln|ln (bN )| = −at + c

Page 1
MAT323(Dynamical Systems) Assignment 2 Ahnaf Shahriar Asif (22301787)

For t = 0 set N = N0 , we get, c = ln|ln (bN0 )|. Putting the value of c,

ln|ln (bN )| = −at + ln|ln (bN0 )|

ln|ln (bN )| − ln|ln (bN0 )| = −at
!
ln (bN )
ln = −at
ln (bN0 )
ln (bN )
= e−at
ln (bN0 )
ln (bN ) = ln (bN0 )e−at
−at
bN = eln (bN0 )e
bN = (bN0 )e−at
1
N = (bN0 )e−at
b
For various initial points, the graph of N (t) is given below:

3.0

2.5

2.0

1.5

1.0

0.5

0.5 1.0 1.5 2.0 2.5

2.3.3

Ṅ
For certain species of organisms, the effective growth rate N is highest at intermediate N .

(a)

Ṅ
pr
Letting N = 0, r = a(N − b)2 =⇒ N = b ± a.

N represents the number of organisms in some species, so it cannot be a negative number. For the
positive sign, it’s already satisfied, but for the negative sign, b > ar .
p

(b)
pr
Letting Ṅ = 0, N (r − a(N − b)2 ) = 0. From that, we get three fixed points, N1 = 0, N2 = b − a,
N3 = b + ar . The graphical representation is given in the next page.
p

Page 2
MAT323(Dynamical Systems) Assignment 2 Ahnaf Shahriar Asif (22301787)

50
Stable
Unstable
40

30

20

10

10

20
0 5 10 15 20 25 30 35 40

(c)

I couldn’t find an analytical solution to the DE. So, I used Sagemath to sketch the trajectories of N (t)
based on different values of N0 .

0.5 1.0 1.5 2.0 2.5

(d)

N
The logistic model equation is Ṅ = rN (1 − K ). I found a few differences between the logistic model and
our current N (t).
• The logistic equation has 2 fixed points, but N (t) in our case has 3 fixed points.
• In both cases, 0 is a fixed point, but in logistic model, 0 is unstable, but here 0 is stable. This
means that the logistic model tells us that the growth of some species can happen even if there is
only one organism! The Alee effect doesn’t do that. It tells us that there needs to be some lower
boundary of organisms that is important for the survival of the organisms.

2.4.8

Ṅ = −aN ln (bN ). N (t) is proportional to the number of tumor cells, and a, b are positive constants. Let
′
f (N ) = Ṅ = −aN ln (bN ). The fixed points will be 0, 1b . Now, the sign of f (N ) at the fixed points will
′ ′
tell us about the stability. if f (N ) > 0, the fixed point will be unstable, and if f (N ) < 0, the fixed point
′ ′
will be stable, and if f (N ) = 0, the stability cannot be determined. Now, f (N ) = −a ln (bN ) − a =
′
−a(ln (bN ) + 1). Now, for N = 1b , f (N ) = −a(ln 1 + 1) = −a < 0, which means N = 1b is a stable point.
′
For N = 0, f (N ) = −a(ln 0 + 1) ≈ ∞ > 0, so N = 0 is an unstable point.

Page 3
MAT323(Dynamical Systems) Assignment 2 Ahnaf Shahriar Asif (22301787)

2.4.9

(a)

ẋ = −x3 . Analytical solution:

dt
ẋ = = x3
dx
−x−3 dx = dt
Z Z
− x−3 dx = dt

x−2
=t−c
2
Applying x(0) = x0 , c = − 2x1 2 . Which gives us 1
2x2 = t+ 1
2x20
=⇒ x(t) = √ ±x0 . Now,
0 1+2tx20
±x0
lim x(t) = ∞ = 0.
t→∞

(b)

dt
ẋ = = −x
Z dx Z
1
− dx = dt
x
ln|x| = −t + ln|c|
x = ce−t
Letting x(0) = x0 , c = x0 . so, x(t) = ce−t . From (a), we got x1 (t) = √ ±x0 . Now, I have plotted both
1+2tx20
of the functions at x0 = 10, up to t = 10. The graph is given below:
10 dx = − x 3
dt
dx = − x
dt

2 4 6 8 10

10

Here, we can see that the blue curve slows down very rapidly than the red one until ±1. Then the red
one decreases faster than the blue one.

3.4.14

ẋ = rx + x3 − x5 . Now, let ẋ = rx + x3 − x5 = 0.

Page 4
MAT323(Dynamical Systems) Assignment 2 Ahnaf Shahriar Asif (22301787)

rx + x3 − x5 = 0
−x(x4 − x2 − r) = 0 =⇒ x1 = 0
or, x4 − x2 − r = 0
y 2 − y − r = 0 Letting y = x2 =⇒ x = ±sqrty
 
√
1 ± 1 + 4r
y=
s 2
√
1 ± 1 + 4r
x=±
2
q √ q √ q √ q √
So, x1 = 0, x2 = 1+ 21+4r , x3 = 1− 21+4r , x4 = − 1+ 21+4r , x5 = − 1− 21+4r . Here, x1 , x2 , x3 , x4 , x5
are the fixed points.

(b)
q
If 1 + 4r = 0, r = − 14 . In that case, we will have only two nonzero fixed points, x = ± 1
2.

If 1 + 4r < 0, r < − 14 . In that case, we won’t have any nonzero fixed points as the roots will be
complex numbers.

If 1 ≤ 1 + 4r, r ≥ 0. In that case, 2 of our fixed points will be complex, so we will ignore them.

Now, I’m plotting the graph for r = 1, 0, − 18 , − 14 , −1.

q
1 + 5 , stable
p

−1, stable
−
100 2
, unstable
0
q 4 0, unstable
1 + 5 , stable
+1, stable
p

50 2

0 0

50
2

4
100

100 50 0 50 100
10 5 0 5 10

Figure 1: r = 1 Figure 2: r = 0

r q

0.6 −
r
1 − 12
2 , stable
q

−
1 + 12
2 , stable
0.4
−1
, half-stable
0.4 0,rstable 2
0, stable
, half-stable
q
+1
, unstable
1
+r 1 −2 2 2
0.2
q

, unstable
1
0.2 + 1 +2 2

0.0 0.0

0.2 0.2

0.4 0.4
0.6 1.0 0.5 0.0 0.5 1.0
0.4 0.2 0.0 0.2 0.4
Figure 4: r = − 41
Figure 3: r = − 18

Page 5
MAT323(Dynamical Systems) Assignment 2 Ahnaf Shahriar Asif (22301787)

0.4
−1
2 , half-stable
0, stable
+1
2 , half-stable
0.2

0.0

0.2

0.4

1.0 0.5 0.0 0.5 1.0

r = −1

(c)

From the above graphs, it is obvious that two saddle-node bifurcations happened at r = − 41 . At this crit-
ical value, two half-stable fixed points collapse and then created two fixed points with opposite stability.

3.4.15

dV (x)
Let f (x) = ẋ = rx + x3 − x5 . Now, dx = −f (x).

dV (x)
= −f (x)
dx Z
V (x) = − f (x)dx
Z
=− rx + x3 − x5
Z Z Z
= − (rx)dx − x3 dx + x5 dx

rx2 x4 x6
=− − +
2 4 6
q √
Now, for fixed points, f (x) = rx+x3 −x5 = 0. From task (3.4.14), we know that x1 = 0, x2 = 1+ 21+4r ,
q √ q √ q √
x3 = 1− 21+4r , x4 = − 1+ 21+4r , x5 = − 1− 21+4r . Here, x1 , x2 , x3 , x4 , x5 are the fixed points.
2 2
Now, the local fixed points for V (x) are dVdx(x) = −f (xmin ) = 0, and d Vdx
(xmin )
2 > 0. Now, d Vdx
(xmin )
2 =
df 2 4
− dx = −r − 3x + 5x . Now, for each fixed point, we have to check if the value is greater than 0.
2
For x = 0, d dx
V (0)
= −r > 0 =⇒ r < 0.
q 2√ 2 √
For x = ± 1+ 21+4r , d dx V (x)
2 = 4r + 4r + 1 + 1 > 0 =⇒ r > − 41 .
q √ 2 √
For x = ± 1− 21+4r , d dx V (x)
2 = 4r − 4r + 1 + 1 > 0 =⇒ r > 0.

Now, 3 fixed points occur at − 41 < r < 0, and 2 occur at r > 0. Since there will be 3 equally deep
local minimums, we will take − 14 < r < 0, which means V (x0 ) = V (x1 ) = V (x2 ). Since V (0) = 0,
v(x1 ) = V (x2 ) = 0. From that, we have to find the value of rc .

Now,

rc x21 x4 x6
V (x1 ) = − − 1+ 1
2 4 6
x21 √ √ 3
= (−8rc − 1 + 4rc − 1) = 0 =⇒ −8rc − 1 + 4rc − 1 = 0 =⇒ rc = −
24 16

Page 6