Introduction to Nonlinear Systems

Daniel A. (MSc. & Science Ambassador)

Department of Electromechanical Engineering

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 1/15

Introduction to Nonlinear and MIMO Systems

Definition
A nonlinear system is a system having output change that is not proportional to the input.
Every nonlinear time-invariant system is represented as

ẋ = f (x, u)
y = h(x, u) (1)

Types of Nonlinearities
• Hard nonlinearity: Includes back-lash, coulombs friction, etc.
• Soft nonlinearity: Nonlinearities due to state and/or input

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 2/15

Determining Equilibrium Points for Nonlinear Dynamic Systems

Key Concept
Nonlinear systems may have a point or set of points in state space where the dynamics of the
system stay there forever after arriving there. These points are called equilibrium points or
limit cycles.

Important Property
At equilibrium point or limit cycles ẋ = 0 always.

Characteristics
• May be single or multiple points
• Some systems may have no equilibrium points
• Stability analysis focuses on these points

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 3/15

Linearization of Nonlinear Systems

Fundamental Approach
A linearization of a nonlinear system is an approximation of the system about some desired
point (operating point). Consider a system:

ẋ = f (x) (2)

with operating point xe .

Why Linearize?
• Enables use of linear analysis tools
• Provides local approximation near operating points
• Foundation for controller design

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 4/15

Small Deviation from Operating Point

Taylor Series Expansion

Assume a small deviation ∆x from the operating point xe :

x = xe + ∆x (3)

The Taylor series expansion about this point:

1 ′′
ẋe + ∆ẋ = f (xe + ∆x) = f (xe ) + f ′ (xe )∆x + f (xe )(∆x)2 + HOT (4)
2!

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 5/15

Neglecting Higher Order Terms

Linear Approximation
Neglecting higher-order terms:

ẋe + ∆ẋ = f (xe ) + f ′ (xe )∆x (5)

Since ẋe = f (xe ):

∆ẋ = f ′ (xe )∆x (6)

Let ∆x = x̃:

x̃˙ = f ′ (xe )x̃ (7)

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 6/15

Deviation as State Variable

Multi-Variable Case
For f (x, u), approximation about (xe , ue ):

∂f ∂f
x̃˙ = x̃ + ũ (8)
∂x xeq ,ueq ∂u xeq ,ueq

System Matrices
∂f ∂f
A= , B=
∂x xeq ,ueq ∂u xeq ,ueq

These define the linearized system matrices.

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 7/15

For nth order system

General Formulation
Consider a nonlinear system with n state dynamics:

ẋi = fi (xi , uj ), i = 1, 2, . . . , n, j = 1, 2, . . . , m

The approximation about an equilibrium point:

∂fi ∂fi
ẋ = x+ u (9)
∂xj xeq ,ueq ∂uj xeq ,ueq

Matrix Definitions
∂fi ∂fi
Aij = , Bij =
∂xj xeq ,ueq ∂uj xeq ,ueq

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 8/15

Introduction to MIMO Systems

Definition
MIMO systems have multiple inputs and multiple outputs. The transfer function consists of
transfer functions from each input to each output:

Yi (S)
= Gij (s) (10)
Ui (s)

where i = 1, 2, 3, . . . , n and j = 1, 2, 3, . . . , m.

Key Features
• Complex interactions between inputs and outputs
• Requires matrix representation
• More challenging analysis than SISO systems

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 9/15

Transfer Function Matrix

Matrix Representation
    
Y1 (s) G11 (s) ··· Gn1 (s) U1 (s)
 ..   .. .. ..   .. 
 . = . . .  .  (11)
Ym (s) G1m (s) · · · Gnm (s) Un (s)

Interpretation
Each Gij (s) represents the transfer function from input j to output i.

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 10/15

Poles of MIMO Systems

Pole Determination
• Poles of individual transfer functions are system poles
• Multiplicity determined by examining minors

Procedure
1. Start with 1 × 1 minors
2. Proceed to larger minors until multiplicity decreases
3. Largest multiplicity found is the pole’s multiplicity

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 11/15

Zeros of MIMO Systems

Complexity
Determining zeros of MIMO systems is more challenging than SISO systems.

Definition
A MIMO system has a zero at z0 if:
• z0 causes rank reduction of the transfer function matrix
• Results in zero rows in the transfer function

Important Note
Zeros of individual transfer functions may not be system zeros.

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 12/15

Example Systems

Example 1: Complex Nonlinear System

ẋ1 = 7x1 + 2 sin x2 − x42 + u31

ẋ2 = ex1 − 3x2 − 1 + 5x21 + sin u2
y1 = ex1 +x2
y2 = x1 + x2

Example 2: Coupled Tank System

√ √
ẋ1 = −a1 x1 + b1 u1 + c12 x2
√ √
ẋ2 = −a2 x2 + b2 u2 + c21 x1
y1 = x1
y2 = x2

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 13/15

Example Systems

Example 3: Complex Nonlinear System

ẋ1 = 3x1 + (x21 + x22 )

ẋ2 = 2x1 x2 (x21 + x22 )
y1 = x1 x2
y2 = x1 + x2

Example 4: Coupled Tank System

ẋ1 = x2
ẋ2 = µ(1 − x21 )x2 − x1
y1 = x1
y2 = x2

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 14/15

Example 5: Competing Species Model

Predator-Prey Dynamics

ẋ1 = αx1 − βx1 x2 + u1

ẋ2 = δx1 x2 − γx2 + u2
y1 = x 1
y2 = x 2

System Interpretation
• States: x1 (prey), x2 (predator) populations
• Inputs: Migration rates u1 , u2
• Key nonlinearity: x1 x2 interaction term

Introduction to Nonlinear Systems Daniel A. (MSc. & Science Ambassador) 15/15