You are on page 1of 51

Digital Control Systems

State-Space Approach
INTRODUCTION
❑ State: The state of a dynamic system is a minimum set of variables, the
knowledge of which completely determines the behavior of the system.

❑State Variable: The variables that comprise the smallest set of variables
termed as the state of the system.

❑State-space: The n-dimensional space whose coordinate axes consist of


the x1-axis, x2-axis,..xn-axis is called a state space.

❑ State-space Equations: We are concerned with three types of variables


that are involved in the modeling of dynamic systems:
▪ input variables,
▪ output variables, and
▪ state variables.
INTRODUCTION
❑ For Linear (or Nonlinear ) discrete-time systems:

❑ For Linear Time-varying discrete-time systems:


INTRODUCTION
❑ For Linear Time-invariant discrete-time systems:

❑ Block diagram representation:


INTRODUCTION
❑ For Linear-Time-Invariant continuous-time systems:
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ Canonical Forms for Discrete-Time State-Space Equations:

❑ Pulse Transfer Function:

❑ There are many ways to realize state-space representations for the discrete time
system represented by these equations:

1. Controllable Canonical Form


2. Observable Canonical Form
3. Diagonal Canonical Form
4. Jordan Canonical Form
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS

❑ Controllable Canonical Form:


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ Controllable Canonical Form:
If we reverse the order of the state variables:
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS

❑ Observable Canonical Form


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
If we reverse the order of the state variables:
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ Diagonal Canonical Form:
If the poles of pulse transfer function are all distinct:
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ Jordan Canonical Form:
If the pulse transfer function involves a multiple pole of order m at z=p1 and all
other poles are distinct:
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ Controllable Canonical Form: Direct Programming Method
❑ Discrete-Time State-Space Equation:

❑ Pulse Transfer Function:


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ Let us define

❑ Re-writing the relation as

one obtains
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ Let us define the state variables as

❑ One clearly has


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ In terms of difference equations, the preceding (n-1) equations become:

❑ Using the state variables in the equation

yields

❑ The corresponding difference equation becomes:


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ One can similarly re-write the equation

as

❑ Which yields the output equation as


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ Re-writing the state equations and output equations in a standard vector-matrix
form one gets Controllable Canonical Form given as
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS

❑ Observable Canonical Form: Nested Programming Method


❑ Re-write the pulse transfer function

as

or
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS

❑ Define State variables as

❑ The last equation may be written in the form


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS

❑ Substituting the relation in the definition of state variables and by multiplying both
sides by z one gets
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS

❑ Taking inverse z-transforms of the preceding n equations and writing them in


reverse order one gets the difference equations

❑ Taking inverse z-transforms of the equation

one gets the output equation


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS

❑ Re-writing the state equations and output equations in a standard vector-matrix


form one gets the Observable Canonical Form given as
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ Diagonal Canonical Form: Partial Fraction Expansion Programming Method
If the poles of pulse transfer function are all distinct:
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS

❑ Define State variables as

❑ The state equations may be written in the form


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS

❑ Similarly one can write

❑ Taking inverse z-transforms of the last 2 equations


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS

❑ Re-writing the state equations and output equations in a standard vector-matrix


form one gets the Diagonal Canonical Form.
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ Jordan Canonical Form: Partial Fraction Expansion Programming Method
If the pulse transfer function involves a multiple pole of order m at z=p1 and all
other poles are distinct:
STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ From the pulse transfer function one obtains

❑ Define the first m State variables as remaining (n-m) state variables as


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ The first m state-variables are related to the next ones as

❑ By taking inverse z-transform one obtains


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS

❑ The output equation is given as

❑ Taking inverse z-transform one obtains


STATE SPACE REPRESENTATIONS OF
DISCRETE-TIME-SYSTEMS
❑ Jordan Canonical Form:
If the pulse transfer function involves a multiple pole of order m at z=p1 and all
other poles are distinct:
NONUNIQUENESS OF STATE SPACE
REPRESENTATIONS
NONUNIQUENESS OF STATE SPACE
REPRESENTATIONS
DISCRETIZATION OF CONTINUOUS-TIME
STATE-SPACE EQUATIONS

❑ In digital control of continuous-time plants one needs to convert continuous-time


state space equations into discrete-time ones.

❑ Such conversion can be done by introducing fictitious samplers and fictitious


holding devices into the continuous-time systems.

❑ The error introduced by discretization may be made negligible by using a


sufficiently small sampling period as compared to the significant time constant of
the system.
DISCRETIZATION OF CONTINUOUS-TIME STATE
SPACE EQUATIONS
❑ The solution to the continuous-time state equation may be
obtained in the following manner.
❑ Now writing the equation in the form

❑ In generalized form:
DISCRETIZATION OF CONTINUOUS-TIME STATE
SPACE EQUATIONS
❑ Assumption:
The input vector changes only at equally spaced sampling instants, i.e., at t= kT.
Consider the continuous-time state equation and output equation

The discrete-time representation will take the form

Note: The matrices G and H depend on sampling period T and once T is fixed they
become constant matrices.
We now assume that the input is sampled and fed to a ZOH so that
DISCRETIZATION OF CONTINUOUS-TIME STATE
SPACE EQUATIONS
DISCRETIZATION OF CONTINUOUS-TIME STATE
SPACE EQUATIONS
DISCRETIZATION OF CONTINUOUS-TIME STATE
SPACE EQUATIONS
DISCRETIZATION OF CONTINUOUS-TIME STATE
SPACE EQUATIONS
TIME RESPONSE BETWEEN TWO CONSECUTIVE
SAMPLING INSTANTS
TIME RESPONSE BETWEEN TWO CONSECUTIVE
SAMPLING INSTANTS
PULSE TRANSFER FUNCTION MATRIX
PULSE TRANSFER FUNCTION MATRIX
PULSE TRANSFER FUNCTION MATRIX
PULSE TRANSFER FUNCTION MATRIX
SOLVING DISCRETE-TIME STATE EQUATIONS
SOLVING DISCRETE-TIME STATE EQUATIONS
STATE TRANSITION MATRIX
Z-TRANSFORM APPROACH FOR SOLVING
DISCRETE-TIME STATE EQUATIONS

You might also like