Scribd
Upload
7 views3 pages

Control Systems Assignment Solutions

Uploaded by

aatmanjn
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
7 views3 pages

Control Systems Assignment Solutions

Uploaded by

aatmanjn
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

EE650 - Basics of Modern Control Systems

Solutions for Assignment - 8


x3
1. i. We have, ẋ1 = x2 , ẋ2 = −x1 + 61 − x2
For finding the equilibrium points,

ẋ1 = 0, ⇒x2 = 0
x3
ẋ2 = 0, ⇒ − x1 + 1 = 0
 2 6 
x1
x1 −1 =0
6
√
x1 = 0, x21 = 6, x1 = ± 6
√ √
Equilibrium points: (0, 0), ( 6, 0), (− 6, 0).
Finding the Jacobian matrix, " #
∂f1 ∂f1
∂x1 ∂x2
J= ∂f2 ∂f2
∂x1 ∂x2
  √
0 1 −1 ± j 3
[J](0,0) = , λ= =⇒ stable node
−1 −1 2
 
√ 0 1
[J]( 6,0) = , λ = −2, 1 =⇒ saddle node
2 −1
 
0 1
[J](−√6,0) = , λ = −2, 1 =⇒ saddle node
2 −1
ii. Similar as the previous part.
2. Already solved in the previous question.
3.
ẋ1 = x2 , ẋ2 = −a sin x1 − bx2

Equilibrium points: −(0, 0), (π, 0).

Consider V = a (1 − cos x1 ) + 0.5x22 . Then differentiating V , we get

V̇ = a sin x1 ẋ1 + x2 ẋ2


V̇ = x2 a sin x1 + x2 (−a sin x1 − bx2 )
V̇ = −bx22

V̇ ≤ 0 =⇒ ∴ System is stable in the sense of Lyapunov.

4. i) f (x) = sin(x) sgn(x)


We know that sin(x) is globally Lipchitz. But, sin(x) sgn(x) is not continuously differentiable at x = 0
because limx→0+ f ′ (x) = 1 ̸= limx→0− f ′ (x) = −1.
Hence, it is locally Lipchitz.
ii) f (x) = tan(x)
Since, the derivative
 is notglobally bounded, it is not globally Lipschitz. It is locally Lipschitz.
−x1 + x1 x2
iii) f (x) =
x2 − x1 x2

1
 
∂f −1 + x2 x1
=
∂x −x2 1 − x1

Since, all the terms are unbounded, it is locally Lipschitz in R2 .


For finding the Lipschitz constant, ∂f
∂x = max {|− |+x2 | + |x1 | , |x2 | + |1 − x1 |} .
∞

∂f
≤ 1 + k1 + k2 = L.
∂x ∞

5. Linearization can be done as shown in Q.1. The phase portraits are made as following:

(a) (i) (b) (ii) (c) (iii)

Figure 1: Phase portraits

6. i) ẋ1 = −x1 + x1 x2 , ẋ2 = −x2

x21 x2
V = + 2
2 2
V̇ = x1 ẋ1 + x2 ẋ2
V̇ = −x21 + x21 x2 − x22

In the set ∥x∥2 ⩽ r2 , we have ∥x1 ∥ < r. Hence,

V̇ ⩽ −x21 − x22 + r |x1 | |x2 |


x2 = e−x2 t x2 (0)
ẋ1 = −1 + e−x2 t x2 (0) x1
 

So, limt→∞ x1 (t) = 0 & limt→∞ x2 (t) = 0.


∴ The system is globally asymptotically stable.
ii)

ẋ1 = −x2 − x1 1 − x21 − x22 , ẋ2 = x1 − x2 1 − x21 − x22


 

x21 x2
v= + 2
2 2
V̇ = x1 ẋ1 + x2 ẋ2
= −x1 x2 − x21 1 − x21 − x22 + x1 x2 − x22 1 − x21 − x22
 

= − x21 + x22 1 − x21 − x22


 

= −2V (1 − 2V )

For a specific region, V is negative definite ⇒ System is locally asymptotically stable.


7. We have,
ẋ(t) = (x − 1)(x + y), ẏ(t) = y − x2
Eq. pts: (1, 1), (0, 0), (−1, 1).
 
2x + y − 1 x − 1
J=
2x 1
 
2 0
J|(1,1) = ⇒ λ = 2, 1
−2 1
 
−1 −1
J|(0,0) = ⇒ λ = −1, 1
0 1
  √
−2 −2 −1 ± j 7
J|(−1,1) = ⇒λ= ⇒ stable node
2 1 2

∴ System is asymptotically stable at (−1, 1).


8. Similar as in Q.3.

Common questions

Powered by AI

The stability of a nonlinear system can be determined using Lyapunov's method by constructing a suitable Lyapunov function that is positive definite and demonstrating that its time derivative along system trajectories is negative definite (or at least non-positive). For instance, in the system \(\dot{x}_1 = -x_2 -x_1(1-x_1^2-x_2^2)\), \(\dot{x}_2 = x_1 -x_2(1-x_1^2-x_2^2)\), the Lyapunov function \( V = \frac{1}{2}(x_1^2 + x_2^2) \) shows \( \dot{V} < 0 \) for the specific region \( \|x\|_2 \leq r^2 \), indicating local stability in that region .

A system is globally asymptotically stable if every solution trajectory tends toward the equilibrium point as time approaches infinity, regardless of initial conditions. In contrast, a system is locally asymptotically stable if only trajectories originating within a certain neighborhood of the equilibrium move towards it over time. Thus, global stability implies stronger robustness to initial conditions compared to local stability .

A stable node is characterized by all eigenvalues of the system's Jacobian matrix at the equilibrium point having negative real parts, indicating all perturbations decay over time and the system returns to equilibrium. Conversely, a saddle node features eigenvalues with both positive and negative real parts, leading to some diverging trajectories, resulting in instability .

When eigenvalues of a Jacobian matrix have both positive and negative parts, it indicates the presence of a saddle node. A saddle node implies that the system is unstable because some trajectories diverge due to the directions associated with positive real parts of the eigenvalues, despite other directions being stable .

The Jacobian matrix evaluated at equilibrium points gives the linear approximation of the system at those points. Analyzing the eigenvalues of the Jacobian matrix helps determine stability: if all real parts of eigenvalues are negative, the point is a stable node; otherwise it is an unstable or saddle node. For the given system, at \((0,0)\), eigenvalues \(\lambda = -1 \pm j\sqrt{3}/2\) indicate stability. At \((\sqrt{6},0)\) and \((-\sqrt{6},0)\), eigenvalues \(\lambda = -2, 1\) indicate saddle nodes, implying instability at these points .

Lyapunov's stability theory is used to show that the system is stable in the sense of Lyapunov by constructing a Lyapunov function \( V = a(1 - \cos x_1) + 0.5x_2^2 \). Differentiating \( V \) along system trajectories gives \( \dot{V} = -bx_2^2 \), which is non-positive, indicating that \( V \) doesn't increase over time, thus confirming stability .

Linearization simplifies a nonlinear system around equilibrium points, providing an approximation that captures the dynamics near those points. By evaluating the Jacobian matrix and examining its eigenvalues for the linearized version, insights are gained into the stability and dynamic responses, such as oscillations or exponential divergence, thus predicting behavior for small perturbations .

Phase portraits graphically represent trajectories of the system in state space, showing how the system behaves near equilibrium points. They help illustrate the stability and nature of equilibria (e.g., nodes, saddles) and provide a visual tool to understand the evolution of system states over time. Different configurations of a phase portrait provide insights into the system's dynamics and how perturbations might evolve .

The equilibrium points are found where both derivatives \( \dot{x}_1 \) and \( \dot{x}_2 \) equal zero. From \( \dot{x}_1 = x_2 = 0 \) and \( \dot{x}_2 = -x_1 + x_3 - \frac{1}{6} \cdot 0 = 0 \), resulting in \( x_1 = 0 \) and \( x_1 = \pm \sqrt{6} \). Thus, the equilibrium points are \( (0, 0) \), \( (\sqrt{6}, 0) \), and \( (-\sqrt{6}, 0) \).

The function \( f(x) = \sin(x) \; \text{sgn}(x) \) is locally Lipschitz because it is not continuously differentiable at \( x = 0 \). The derivatives from the positive and negative sides do not match: \( \lim_{x \to 0^+} f'(x) = 1 \neq \lim_{x \to 0^-} f'(x) = -1 \). This discontinuity prevents it from being globally Lipschitz .

You might also like