9 visualizações

Enviado por Alejandro Soto Altamirano

- 05 Nonlinear Dynamics
- Active Vibration Control of a Smart Structure
- AspenHYSYSDynamicModeling
- ICRAMID_120
- CE 2000 GATE Question Paper
- Detection of Bias, Drift, Freeze and Abrupt Sensor Failure using Intelligent Dedicated Observer Based Fault Detection and Isolation for Three Interacting Tank Process
- CATIA-datasheet-v22
- Strozzi Et Al-1999-AIChE Journal
- Assignment 1 Sem21718
- 0. SYLLL
- PIEAS M.S Fellowship Interview
- Nonlinear Dinamical and Control Systems
- Automatics i Calcula
- Roll 8 Emplicit Implicit Transient Dynamic Analysis (Pradeep Shrestha)
- Matlab Intro
- linearization
- Chap5
- A Comparative Analysis of PID, Fuzzy and H Infinity Controllers Applied to a Stewart Platform
- Nonlinear autonomous control of a Two-wheeled inverted pendulum mobile robot based on sliding mode
- 16CI622_CoursePlan

Você está na página 1de 16

June 2009

Summary: This document follows the lectures on feedback linearization tought at the

University of Zagreb, Faculty of Electrical Engineering and Computing.

1 Motivation

The dynamic equation that describes the water level h(t) in a tank with an opening at the bottom

(cross section is a) is shown with (1). The assumption is that the tank has the same cross section

A throughout the whole height. Input water ow is denoted with q(t).

h(t) =

1

A

q (t)

a

A

_

2gh(t) (1)

The aim is not to linearize the system, which would give approximate dynamics around the

operating point, but we want to design a controller such that it linearizes the system completely, in

the whole operating region. We can resolve this by introducing a new control input v such that

q(t) = a

_

2gh(t) + Av(t).

The system dynamics can now be written as

h(t) = v(t).

Since the systemis completely linear, one can design a controller so that the system is guided to

the desired value, e.g. by choosing v(t) = K [h

des

(t) h(t)] the control closed loop differential

equation becomes

h(t) = K [h

des

(t) h(t)]

1

K

h(t) + h(t) = h

des

(t)

h(s)

h

des

(s)

=

1

1

K

s+1

or h(t) h

des

(t) as t . The following part of this lecture will describe how to formalize the

described approach, what to do with unstable dynamics, etc.

1

2 Introduction

In this part a short description of Lie algebra on vector elds is described, together with an idea of

relative degree in nonlinear systems.

2.1 Lie Algebra

Denition 1 (Lie derivatives). Let h : R

n

R be a smooth scalar function and f : R

n

R be a

smooth vector eld on R

n

, then the Lie derivative of h with respect to f is a scalar function dened

by L

f

h = hf.

The symbol is often written as

x

making the denition as follows: L

f

h =

h

x

f. The Lie

derivative is interpreted as differentiation of function h in the direction of the vector f.

Example 2 (Lie derivative). If a homogenous nonlinear system is described as

x = f(x)

y = h(x)

then the derivatives of the output can be represented using the Lie derivatives:

y =

h

x

x = L

f

h

y =

[L

f

h]

x

x = L

f

2

h

.

Denition 3 (Lie brackets). Let f and g be two vector elds on R

n

. The Lie bracket of f and g is

a third vector dened by

[f, g] = g f f g

.

The Lie bracket [f, g] is commonly written as ad

f

g where ad stands for adjoint. Repeated

Lie brackets can be dened recursively by

ad

f

o

g = g

ad

f

i

g = [f, ad

f

i1

g] for i = 1, 2, ....

.

Some properties of Lie brackets:

bilinearity

[

1

f

1

+

2

f

2

, g] =

1

[f

1

, g] +

2

[f

2

, g]

[f,

1

g

1

+

2

g

2

] =

1

[f, g

1

] +

2

[f, g

2

]

skew comutativity

[f, g] = [g, f]

2

Jacobi identity

L

ad

f

g

h = L

f

L

g

h L

g

L

f

h.

Example 4 (Lie brackets). Let x = f(x) + g(x)u with two vector elds f and g dened by

f =

_

2x

1

+ ax

2

+ sin x

1

x

2

cos x

1

_

g(x) =

_

0

cos(2x

1

)

_

The Lie bracket can be computed as

[f, g] =

_

0 0

2 sin(x

1

) 0

_ _

2x

1

+ ax

2

+ sin(x

1

)

x

2

cos x

1

_

_

2 + cos x

1

a

x

2

sin x

1

cos x

1

_ _

0

cos(2x

1

)

_

=

_

a cos(2x

1

)

cos x

1

cos(2x

1

) 2 sin(2x

1

)(2x

1

+ ax

2

+ sin x

1

)

_

.

2.2 Relative Degree

A SISO nonlinear system is given in a form (2).

x = f(x) + g(x)u

y = h(x)

(2)

If we observe the output derivatives we obtain the following:

y =

h

x

x =

=

h

x

[f(x) + g(x)u] =

=

h

x

f(x)

. .

L

f

h

+

h

x

g(x)

. .

Lgh

u =

= L

f

h + (L

g

h) u

Lets assume that the rst derivative of the output is not inuenced by the input signal u, i.e.

L

g

h = 0 and y = L

f

h. Then the next derivative of the output is

y =

d

dt

y =

d

dt

(L

f

h) =

(L

f

h)

x

x =

=

L

f

h

x

(f + gu) = L

f

(L

f

h) + L

g

(L

f

h) u =

= L

2

f

h + L

g

(L

f

h)

. .

=0

u

Induction leads us to the following:

y = L

f

h

y = L

2

f

h

.

.

.

y

(r)

= L

r

f

h + L

g

_

L

r1

f

h

_

u

Lets assume that the r

th

derivation of the output sees the input signal u. Then the value r is

called the relative degree of the system.

3

Denition 5 (Relative degree). The nonlinear system

x = f(x) + g(x)u

y = h(x)

has relative degree r at the point x

0

if

L

g

_

L

k

f

h(x)

0

and all k < r 1

L

g

_

L

r1

f

h(x

0

)

= 0.

In other words, we should differentiate the output until the input appears in the equation. The

derivative to which we have come is the degree of the system.

Note: Relative degree is dened only if there exists an output from the system. It

also depends on the chosen output equation. In linear systems the relative degree is

the pole excess of the system, i.e. the difference between the number of poles and the

number of nite zeros of the transfer function of the systems. If the relative degree of

the linear system is less then n, that means that there exist nite zeros (which may be

unstable). Since transfer functions do not exist in nonlinear systems, the fact that the

relative degree of the nonlinear system is smaller then the order of the system, means

that there are some internal dynamics which cannot be seen from the output.

Example 6 (Relative degree of the Van der Pool oscillator). Calculate the relative degree of the

van der Pool oscillator given with

x = f(x) + g(x)u =

_

x

2

2 (1 x

2

1

) x

2

2

x

1

_

+

_

0

1

_

u

y = h(x) = x

1

For starters, lets differentiate the output

L

g

L

0

f

h = L

g

h =

h

x

g =

_

h

x

1

h

x

2

_

0

1

_

=

_

1 0

_

0

1

_

= 0

This one does not depend on the input!!

L

g

L

1

f

h = L

g

_

h

x

f

= L

g

_

_

1 0

_

x

2

2 (1 x

2

1

) x

2

2

x

1

__

= L

g

x

2

=

x

2

x

g =

=

_

0 1

_

0

1

_

= 1

This one depends on the input so the relative degree is r = 2 (we had to make 2 calculations, i.e.

we had to differentiate the output twice to get to the input signal).

4

If the output were dened as:

y = h(x) = sin (x

2

) ,

then the following calculations would take place

L

g

L

0

f

h = L

g

h =

h

x

g =

_

h

x

1

h

x

2

_

0

1

_

=

_

0 cos (x

1

)

_

0

1

_

= cos (x

1

)

so the relative degree is 1 at all points apart from x =

2

(2k + 1). At these points, relative degree is

not dened. From this example it is obvious that the relative degree depends on the chosen output,

just as the number of zeros in the linear system depend on the chosen output.

3 Input/Output Linearization

We have shown before that the nonlinear system (2) can be transformed to a form

y = L

0

f

h

y = L

1

f

h

y = L

2

f

h

.

.

.

y

(r)

= L

r

f

h + L

g

_

L

r1

f

h

_

u.

If we introduce (x) = L

r

f

h and (x) = L

g

_

L

r1

f

h

_

then y

(r)

= (x) + (x) u. Since the

input signal can be whatever we want it to be, lets choose it in such a way that the nonlinearities

in the system are compensated for, i.e. y

(r)

= (x) + (x) u v from where it follows that

v = (x) + (x) u

u =

1

(x)

[(x) + v]

Graphically, this is what happened:

Another way of thinking is that we have dened a new state space vector, which consists of the

derivatives of the output:

z

1

= y = L

0

f

h

z

2

= y = L

1

f

h

.

.

.

z

r

= y

(r1)

= L

r1

f

h

z

r+1

= y

(r)

= + u = v

5

So the system can be represented in a state space form as

_

_

z

1

z

2

.

.

.

z

r

_

_

=

_

_

0 1 0

0 0

.

.

. 0

0 0 1

0 0 0

_

_

_

_

z

1

z

2

.

.

.

z

r

_

_

+

_

_

0

0

.

.

.

1

_

_

v

y =

_

1 0 0

_

z

1

z

2

.

.

.

z

r

_

_

which is in fact a cascade of r integrators. Now we can design any type of controller for the r

integrator linear system. The most appropriate one would be a state controller. What was the

transformation between the two state spaces, x and z?

z = (x) =

_

1

(x)

2

(x)

.

.

.

r

(x)

_

_

=

_

_

L

0

f

h

L

1

f

h

.

.

.

L

r1

f

h

_

_

Algorithm for input/output linearization:

1. Find the relative degree r of the nonlinear system.

L

g

_

L

k

f

h(x)

= 0

L

g

_

L

r1

f

h(x

0

)

= 0

2. Make the r state transformations.

z = (x) =

_

1

(x)

2

(x)

.

.

.

r

(x)

_

_

=

_

_

L

0

f

h

L

1

f

h

.

.

.

L

r1

f

h

_

_

3. Dene the new input so that the states are linearized.

(x) = L

r

f

h

(x) = L

g

_

L

r1

f

h

_

u =

1

(x)

[(x) + v]

6

Example 7 (r = n).

x =

_

_

0

x

1

+ x

2

2

x

1

x

2

_

_

+

_

_

e

x

2

e

x

2

0

_

_

u

y = h(x) = x

3

Find the relative degree

L

g

L

0

f

h = L

g

h =

h

x

g =

=

_

h

x

1

h

x

2

h

x

3

_

_

e

x

2

e

x

2

0

_

_

= 0

L

g

L

1

f

h = L

g

_

h

x

f

= L

g

_

_

_

0 0 1

_

_

0

x

1

+ x

2

2

x

1

x

2

_

_

_

_

=

= L

g

[x

1

x

2

] =

_

1 1 0

g =

_

1 1 0

_

_

e

x

2

e

x

2

0

_

_

= 0

L

g

L

2

f

h = L

g

L

f

_

h

x

f

= L

g

L

f

[x

1

x

2

] = L

g

_

_

_

_

1 1 0

_

_

0

x

1

+ x

2

2

x

1

x

2

_

_

_

_

_

=

= L

g

[x

1

x

2

2

] =

_

1 2x

2

0

_

_

e

x

2

e

x

2

0

_

_

= e

x

2

(1 + 2x

2

)

Relative degree is 3 if 1 + 2x

2

= 0, it is the same as the order of the nonlinear system.

Make the r state transformations.

z

1

=

1

(x) = L

0

f

h = h = x

3

z

2

=

2

(x) = L

1

f

h =

_

0 0 1

_

_

0

x

1

+ x

2

2

x

1

x

2

_

_

= x

1

x

2

z

3

=

3

(x) = L

2

f

h = L

f

L

f

h =

_

1 1 0

_

_

0

x

1

+ x

2

2

x

1

x

2

_

_

= x

1

x

2

2

Dene the new input so that the states are linearized.

(x) = L

3

f

h = L

f

L

2

f

= L

f

_

x

1

x

2

2

=

_

1 2x

2

0

_

_

0

x

1

+ x

2

2

x

1

x

2

_

_

= 2x

2

_

x

1

+ x

2

2

_

(x) = L

g

_

L

2

f

h

_

= L

g

_

x

1

x

2

2

=

_

1 2x

2

0

_

_

e

x

2

e

x

2

0

_

_

= (1 + 2x

2

) e

x

2

7

u =

1

(x)

[(x) + v] =

2x

2

(x

1

+ x

2

2

)

(1 + 2x

2

) e

x

2

1

(1 + 2x

2

) e

x

2

v

By introducing this control, we have obtained a system with the following dynamics

_

_

z

1

z

2

z

3

_

_

=

_

_

0 1 0

0 0 1

0 0 0

_

_

_

_

z

1

z

2

z

3

_

_

+

_

_

0

0

1

_

_

v

y =

_

1 0 0

_

_

z

1

z

2

z

3

_

_

and now we can design whatever type of control we desire.

Lets take a look at an example where relative degree is smaller than the order of the system

(rn).

Example 8 (r < n THIS EXAMPLE IS WRONG REGARDING STABILITY - NEEDS REVI-

SION).

x

1

= x

2

1

x

2

x

2

= 3x

2

+ u

y = h(x) = 2x

1

x

2

The state-space form is then

_

x

1

x

2

_

=

_

x

2

1

x

2

3x

2

_

+

_

0

1

_

u

y = h(x) = 2x

1

x

2

Find the relative degree

L

g

L

0

f

h = L

g

h =

h

x

g =

_

h

x

1

h

x

2

_

0

1

_

=

_

2 1

_

0

1

_

= 1

Relative degree is 1. That means that we have 1 state transformation. Notice that we have a

second order system!!

Make the r state transformations.

z

1

=

1

(x) = L

0

f

h = h = 2x

1

x

2

The second state stays the same.

Dene the new input so that the states are linearized.

(x) = L

1

f

h =

h

x

f =

_

2 1

f = 2x

2

1

x

2

3x

2

8

(x) = L

g

_

L

0

f

h

_

= L

g

h = 1

u =

1

(x)

[(x) + v]

Now the system of nonlinear equations is transformed to

z

1

= v

x

2

= 3x

2

+ u = 3x

2

2x

2

1

x

2

3x

2

v = 2x

2

1

x

2

v

Lets observe the zero dynamics. This is done so that the output is kept 0 and the behavior of

the system is observed. If the output is zero y = 2x

1

x

2

= 0 x

2

= 2x

1

which means that

z

1

= v

x

2

= 2x

3

2

v

what is a stable system. This means that the zero dynamics are stable.

Lets take a look at another example where the zero dynamics are not stable.

Example 9 (r < n, unstable zero dynamics).

x

1

= x

2

+ x

3

1

+ u

x

2

= u

y = h(x) = x

1

The state space form is then

_

x

1

x

2

_

=

_

x

2

+ x

3

1

0

_

+

_

1

1

_

u

y = h(x) = x

1

Find the relative degree

L

g

L

0

f

h = L

g

h =

h

x

g =

_

h

x

1

h

x

2

_

1

1

_

=

_

1 0

_

1

1

_

= 1

Relative degree is 1. That means that we have 1 state transformation. Notice that we have a second

order system!!

Make the r state stransformations

z

1

=

1

(x) = L

0

f

h = h = x

1

The second state stays the same. Dene the new input so that the states are linearized.

(x) = L

1

f

h =

h

x

f =

_

1 0

_

x

2

+ x

3

1

0

_

= x

2

+ x

3

1

9

(x) = L

g

_

L

0

f

h

_

= L

g

h = 1

u =

1

(x)

[(x) + v] = x

2

x

3

1

+ v

Now the system of nonlinear equations is transformed as

z

1

= v

x

2

= x

2

+ x

3

1

v

Not only is it not linearized, but lets say that we want to design a proportional controller so

that the output goes to 0, i.e. v = K

p

(y y

des

) = K

p

y The closed loop system is then

z

1

= v = K

p

z

1

x

2

= x

2

+ z

3

1

+ K

p

z

1

If y has to go to 0, that means that z

1

has to go to zero. If that one goes to zero, that means that

x

2

is governed by x

2

= x

2

which is unstable!!!!!

Obviously we have some internal dynamics which are unstable. The reason why we didnt

notice these internal dynamics is the fact that the relative degree is smaller than the degree of the

nonlinear system. There is something else, some other states that we forgot about - n r states.

The internal dynamics might not be unstable, maybe it would be enough for us to linearize the r

states of the system (which are based on the output) and the rest would have been stable. But we

cannot be sure. Bare in mind that we performed only the transformation to the r states.

The question is: can we derive such a transformation so that the relative degree of the system

is n? This way we could include all of the dynamics in the design procedure. What should this

transformation look like? What are the constraints? Is this transformation even possible? What if

the relative degree of the system is not equal to the degree of the nonlinear system, i.e. rn? This

means that we have the following situation

_

_

z

1

.

.

.

z

r

r+1

.

.

.

n

_

_

=

_

(x)

(x)

_

=

_

_

L

0

f

h(x)

.

.

.

L

r1

f

h(x)

r+1

(x)

.

.

.

n

(x)

_

_

4 Input/State Linearization

Lets rephrase the problem. If a nonlinear system is given with

x = f(x) + g(x)u

Find a transformation = (x) so that the system in the new coordinates can be linearized and

so that it is controllable. can be linearized. This will be possible if we nd an output y = (x)

such that the system has relative degree n at some x

0

. How to nd a suitable (x)? Just as in the

case when the output was available, this function has to satisfy the following properties:

10

L

g

(x) = L

g

L

f

(x) = ... = L

g

L

n2

f

(x) = 0 (all the terms next to the input u have to

disappear)

L

g

L

n1

f

(x

0

) = 0

.

It can be shown by the use of the Jacobi identity that these conditions can be rewritten as

(x) g = (x) ad

f

g = ... = (x) ad

n2

f

g = 0

(x) ad

n1

f

g = 0

or

(x)

_

g ad

f

g ad

n2

f

g

= 0

(x) ad

n1

f

g = 0.

These equations are much more appropriate since the unknown function (x) has been sepa-

rated from the rest of the calculations.

If we manage to nd this function, the state transformation will follow as

_

1

.

.

.

n

_

_

= (x) =

_

1

(x)

.

.

.

n

(x)

_

_

=

_

_

L

0

f

(x)

.

.

.

L

n1

f

(x)

_

_

How do we know that there even exists a transformation such that the relative degree is n? It is

the same as asking: how do we know that the previously stated partial differential equations have

a solution? The Frobenius theorem tells us if the solution exists.

Theorem 10 (Frobenius). The nonlinear system with f(x) and g(x) being smooth vector elds, is

input-state linearized if and only if there exists a region such that the following conditions hold:

1. the vector elds {g, ad

f

g, ..., ad

n1

f

g} are linearly independent in (the matrix has to have

full rank)

2. the set {g, ad

f

g, ..., ad

n2

f

g} is involutive in ,

Note: The rst condition can be interpreted as controllability condition for the nonlin-

ear system. The involutivity condition is less intuitive. It is trivially satised for linear

systems, but not generally satised in the nonlinear case.

11

Denition 11 (Involutivity). A linear independent set of vectors [f

1

, ..., f

m

] is involutive if

[f

i

, f

j

] =

m

k=1

ijk

(x) f

k

(x),

i.e. if you take Lie brackets you dont generate new vectors.

The input-state linearization can be performed through the following steps:

1. Construct the vector elds g, ad

f

g, ..., ad

n1

f

g for the given system.

2. Check whether the controllability and involutivity conditions are satised.

3. If both are satised, nd the output function (x) (the output function which will lead to

the input-output linearization of the relative degree n) by using

(x)

_

g ad

f

g ad

n2

f

g

= 0

(x) ad

n1

f

g = 0

4. Compute the state transformation

_

1

.

.

.

n

_

_

= (x) =

_

1

(x)

.

.

.

n

(x)

_

_

=

_

_

L

0

f

(x)

.

.

.

L

n1

f

(x)

_

_

and the input transformation with

(x) = L

n

f

(x)

(x) = L

g

_

L

n1

f

(x)

_

u =

1

(x)

[(x) + v]

.

Example 12. Consider a mechanism given by the dynamics which represents a single link exible

joint robot. Its equations of motion is derived as

I q

1

+ MgLsin q

1

+ k(q

1

q

2

) = 0

J q

2

k(q

1

q

2

) = u

Because nonlinearities (due to gravitational torques) appear in the rst equation, while the control

input u enters only in the second equation, there is no easy way to design a large range controller.

x =

_

_

q

1

q

1

q

2

q

2

_

_

, f =

_

_

x

2

MgL

l

sin x

1

k

l

(x

1

x

3

)

x

4

k

J

(x

1

x

3

)

_

_

, g =

_

_

0

0

0

1

J

_

_

12

ad

f

g = [f, g] =

g

x

f

f

x

g =

=

_

_

0 1 0 0

MgL

I

cos x

1

k

I

0

k

I

0

0 0 0 1

k

J

0

k

J

0

_

_

_

_

0

0

0

1

J

_

_

=

_

_

0

0

1

J

0

_

_

ad

2

f

g = [f, ad

f

g] =

ad

f

g

x

f

f

x

ad

f

g =

=

_

_

0 1 0 0

MgL

I

cos x

1

k

I

0

k

I

0

0 0 0 1

k

J

0

k

J

0

_

_

_

_

0

0

1

J

0

_

_

=

_

_

0

k

JI

0

k

J

2

_

_

ad

3

f

g =

_

f, ad

2

f

g

=

ad

2

f

g

x

f

f

x

ad

2

f

g =

=

_

_

0 1 0 0

MgL

I

cos x

1

k

I

0

k

I

0

0 0 0 1

k

J

0

k

J

0

_

_

_

_

0

k

JI

0

k

J

2

_

_

=

_

k

JI

0

k

J

2

0

_

_

[g ad

f

g ad

2

f

g ad

3

f

g] =

_

_

0 0 0

k

IJ

0 0

k

IJ

0

0

1

J

0

k

J

2

1

J

0

k

J

2

0

_

_

It has rank 4. Furthermore, since the above vector elds are constant, they form an involutive

set. Therefore the system is input-state linearizable.

Let us nd the state-transformation z = (x) and the input transformation u = (x)+(x)v

so that input-state linearization is achieved.

(x) g =

_

x

1

x

2

x

3

x

4

g = 0

x

4

= 0

(x) ad

f

g =

_

x

1

x

2

x

3

x

4

_

0

0

1

J

0

_

_

= 0

x

3

= 0

(x) ad

2

f

g =

_

x

1

x

2

x

3

x

4

_

0

k

IJ

0

k

J

2

_

_

= 0

x

2

= 0

(x) ad

3

f

g =

_

x

1

x

2

x

3

x

4

k

IJ

0

k

J

2

0

_

_

= 0

x

1

= 0

13

It is obvious that lambda should be a function of x

1

only. Therefore, we choose a function

which is a diffeomorphism, e.g. (x) = x

1

.

The other states are obtained from the following calculations:

z

2

= L

f

= z

1

f = x

2

z

3

= L

2

f

= z

2

f =

MgL

I

sin x

1

k

I

(x

1

x

3

)

z

4

= L

3

f

= z

3

f =

MgL

I

x

2

cos x

1

k

I

(x

2

x

4

)

The input transformation is then

(x) = L

f

4

(x) = L

f

z

4

=

=

_

MgL

I

x

2

sin x

1

MgL

I

cos x

1

k

I

0

k

I

_

x

2

MgL

I

sin x

1

k

I

(x

1

x

3

)

x

4

k

J

(x

1

x

3

)

_

_

=

=

MgL

I

x

2

2

sin x

1

_

MgL

I

cos x

1

k

I

_ _

MgL

I

sin x

1

+

k

I

(x

1

x

3

)

_

+

k

2

IJ

(x

1

x

3

)

(x) = L

g

L

f

3

(x) = L

g

z

4

=

_

MgL

I

x

2

sin x

1

MgL

I

cos x

1

k

I

0

k

I

_

0

0

0

1

J

_

_

=

k

IJ

with

u =

1

(v )

We end up with the following set of linear equations

z

1

= z

2

z

2

= z

3

z

3

= z

4

z

4

= v

thus completing the input/output linearization.

Example 13.

x =

_

_

x

3

(1 + x

2

)

x

1

x

2

(1 + x

1

)

_

_

+

_

_

0

1 + x

2

x

3

_

_

u

First to see if it is possible to nd a function so that the system can be input/state linearized.

We have to nd [g ad

f

g ad

2

f

g].

14

ad

f

g =

g

x

f

f

x

g =

=

_

_

0 0 0

0 1 0

0 0 1

_

_

_

_

x

3

(1 + x

2

)

x

1

x

2

(1 + x

1

)

_

_

_

_

0 x

3

1 + x

2

1 0 0

x

2

1 + x

1

0

_

_

_

_

0

1 + x

2

x

3

_

_

=

=

_

_

0

x

1

x

2

(1 + x

1

)

_

_

_

_

0

0

(1 + x

2

) (1 + x

1

)

_

_

=

_

_

0

x

1

(1 + x

1

) (1 + 2x

2

)

_

_

ad

2

f

g = [f, ad

f

g] =

(ad

f

g)

x

f

f

x

(ad

f

g) =

=

_

_

0 0 0

1 0 0

(1 + 2x

2

) 2 (1 + x

1

) 0

_

_

_

_

x

3

(1 + x

2

)

x

1

x

2

(1 + x

1

)

_

_

_

_

0 x

3

1 + x

2

1 0 0

x

2

1 + x

1

0

_

_

_

_

0

x

1

(1 + x

1

) (1 + 2x

2

)

_

_

=

=

_

_

0

x

3

(1 + x

2

)

x

3

(1 + x

2

) (1 + 2x

2

) 2x

1

(1 + x

1

)

_

_

_

_

x

1

x

3

(1 + x

2

) (1 + x

1

) (1 + 2x

2

)

0

x

1

(1 + x

1

)

_

_

=

=

_

_

x

1

x

3

+ (1 + x

2

) (1 + x

1

) (1 + 2x

2

)

x

3

(1 + x

2

)

x

3

(1 + x

2

) (1 + 2x

2

) 3x

1

(1 + x

1

)

_

_

Lets observe the necessary conditions around the equilibrium point x = 0.

_

g ad

f

g ad

2

f

g

x

0

=

_

_

0 0 1

1 0 0

0 1 0

_

_

The rank is full, so the rst condition is satised. The second condition is involutivity.

[g, ad

f

g] =

ad

f

g

x

g

g

x

ad

f

g =

=

_

_

0 0 0

1 0 0

(1 + 2x

2

) 2 (1 + x

1

) 0

_

_

_

_

0

1 + x

2

x

3

_

_

_

_

0 0 0

0 1 0

0 0 1

_

_

_

_

0

x

1

(1 + x

1

) (1 + 2x

2

)

_

_

=

=

_

_

0

0

2 (1 + x

1

) (1 + x

2

)

_

_

_

_

0

x

1

(1 + x

1

) (1 + 2x

2

)

_

_

=

=

_

_

0

x

1

3 (1 + x

1

) (1 + 2x

2

)

_

_

15

The matrix

_

g ad

f

g [g, ad

f

g]

has rank 2 which means that the vectors that for it are

linearly dependant. That means that the third column is not a new vector but a linear combination

from the rst two columns. This means that the desired distribution is involutive.

The desired function (x) is then calculated from

(x)

_

g ad

f

g ad

n2

f

g

= 0

_

x

1

x

2

x

3

_

_

0

1 + x

2

x

3

0

x

1

(1 + x

1

) (1 + 2x

2

)

_

_

= 0

and

(x) ad

n1

f

g = 0.

This can be done with (x) = x

1

which is a diffeomorphism.

16

- 05 Nonlinear DynamicsEnviado porrazvan.curtean5992
- Active Vibration Control of a Smart StructureEnviado porHemantha Kumar R
- AspenHYSYSDynamicModelingEnviado porManish Jaiswal
- ICRAMID_120Enviado porSundaraPandiyan
- CE 2000 GATE Question PaperEnviado porPrashant Sunagar
- CATIA-datasheet-v22Enviado porseventhhemanth
- Strozzi Et Al-1999-AIChE JournalEnviado porer
- 0. SYLLLEnviado porNishant Kamal
- Detection of Bias, Drift, Freeze and Abrupt Sensor Failure using Intelligent Dedicated Observer Based Fault Detection and Isolation for Three Interacting Tank ProcessEnviado porijcsn
- Assignment 1 Sem21718Enviado porChoy Beiyee
- PIEAS M.S Fellowship InterviewEnviado porMuhammad Yar Saqib
- Nonlinear Dinamical and Control SystemsEnviado porsable14
- Automatics i CalculaEnviado porMosher Jovan
- Roll 8 Emplicit Implicit Transient Dynamic Analysis (Pradeep Shrestha)Enviado porAsim Gautam
- Matlab IntroEnviado porldy8999
- linearizationEnviado porhraiserdias
- Chap5Enviado poreshbli
- A Comparative Analysis of PID, Fuzzy and H Infinity Controllers Applied to a Stewart PlatformEnviado porIJAERS JOURNAL
- Nonlinear autonomous control of a Two-wheeled inverted pendulum mobile robot based on sliding modeEnviado porAnuradha Ranasinghe
- 16CI622_CoursePlanEnviado porRamaDinakaran
- Analyzing Control Systems with DelaysEnviado porMuhammad Ali Masood Cheema
- 1-s2.0-S0957415814001111-mainEnviado porVictor Daniel Waas
- IE301 03 Review Linear AlgebraEnviado porf309356
- Solution Part1Enviado porrdehghan66
- 2006s.pdfEnviado porGanesh Kumar
- 1-s2.0-S0020746212001382-mainEnviado pornoured547244
- Full TextEnviado porThinh Vo
- Bootcamp 2011 AlgebraEnviado porpetere056
- 26-1-3Enviado porLic Walter Andrés Ortiz Vargas
- nonlinear Dynamic courseEnviado porAlireza Azerila

- 322192a + PID Control Toolset User ManualEnviado porElvis Yu
- Simocode - Rele InteligenteEnviado porAlejandro Soto Altamirano
- PA_eRTU_Enviado porAlejandro Soto Altamirano
- Catálogo Sicam Pas EnEnviado porAlejandro Soto Altamirano
- Presentazione GPIB- PXI- VXI -Parte IIEnviado porAlejandro Soto Altamirano
- Presentazione GPIB- PXI- VXI -Parte IEnviado porAlejandro Soto Altamirano
- 06 - Digital IOEnviado porAlejandro Soto Altamirano
- 04 - Analog InputEnviado porAlejandro Soto Altamirano
- 03 - Analog and Digital TriggeringEnviado porAlejandro Soto Altamirano
- 02 - Data Acquisition HW - SWEnviado porAlejandro Soto Altamirano
- 01 - Overview of SignalsEnviado porAlejandro Soto Altamirano
- 07 - CountersEnviado porAlejandro Soto Altamirano
- FtpEnviado porAlejandro Soto Altamirano
- Tutorial. 4 CONTROL DESIGNEnviado porSteve Goke Ayeni
- 10580-00178BEnviado porAlejandro Soto Altamirano
- 11410-00583CEnviado porAlejandro Soto Altamirano
- MN240 MI INEnviado porAlejandro Soto Altamirano
- Pc1150 Mi IngEnviado porAlejandro Soto Altamirano
- Fsa210 Mi InEnviado porAlejandro Soto Altamirano
- chemical process control with matlabEnviado porluisagarcia366
- Manual Ejercicios Seminario Daq y LabviewEnviado porAlejandro Soto Altamirano
- Control and Simulation in LabVIEWEnviado porGordana Janevska
- Ecee4638manual OLDEnviado porDavidleonardo Galindo
- 89762720 Manual de Ejercicios MultisimEnviado porRudi Ponce Aguilar
- L01-Intro_to_LogixEnviado porAlejandro Soto Altamirano

- Modeling and Analysis of Dynamic SystemsEnviado porThe-Hanh Pham
- Vlsi and Embedded Systems 05 Ec60xxEnviado porwhiteelephant93
- ECE 333-SyllabusEnviado porCisco Gama
- dsp viva questions.pdfEnviado porMallikharjuna Goru
- MidEnviado porsanthie
- Signals and Systems With SolutionsEnviado porBikram Paul
- Modeling in the Frequency DomainEnviado porAbril de Vera
- Kretchmar Nnhvac Papers Matt DissEnviado porbacuoc.nguyen356
- ch2Enviado porBenyVazquez
- Response of First Order System in Various InputEnviado porDonna Joy Mallari
- Comm-02-Fourier Theory and Communication SignalsEnviado porArquimedes Paschoal
- Assignment 2Enviado porLeon Kok Seng
- System Identification With Matlab. Linear ModelsEnviado porkrishnanand
- Proakis ProblemsEnviado porJoonsung Lee
- 1. Introduction to Control SystemEnviado pormarina890416
- Lecture 1Enviado porBrij Mohan Singh
- MCDS Exercise 4Enviado porberkanhzrc67
- Lecture 5 - Convolution.pdfEnviado porTausique Sheikh
- Takagi-Sugeno-Kang Fuzzy Structures in Dynamic System ModelingEnviado porrafa_mas
- Linearity and Non LinearityEnviado porAdnan Mir
- Overview Random ProcessesEnviado porOzlem Ozgun
- Chap7 Random ProcessEnviado porenyoya
- Nonlinear Control Systems.Enviado porronhill
- 201869_Chapter 2A SlidesEnviado porYuHua
- Embedded Control Systems in CC++ - An Introduction for Software Developers Using MATLABEnviado pordiminano2745
- Time Domain Analysis of SignalsEnviado porshashankif
- newEE235notes.pdfEnviado porYara Sherif
- BME532courseoutlineFall13(2).pdfEnviado porandrewchristie07
- EE202 2017 Chapter1 Part2 Mar15Enviado por김승희
- 5 - Properties of LTI SystemsEnviado porspacesquirrel