ROBOT CONTROL

Chapter 8
 What is a Control System?
• An interconnection of components
forming a system to provide a
desired system response
• The study of control provides us with
a process for analyzing and
understanding the behavior of a
system with some input
• It also introduces methods for
achieving the desired system
response
 What sorts of Systems?
Example: systems which uses control system
techniques
– Vehicle Control and Design (land, sea, air, space)
– understanding how the system responds to
external disturbances
– Biomedical (cardiac system, dialysis machine) –
design and control of the system with relation to
human interaction
– Stock Market - understanding the relationship
between interest rates and inflation where one
drives the other
– Biological - determining the relationship between
the outcome of a natural phenomenon when the
system is disturbed
 Ex: Flyball Governor
Flyball Governor (1788)
• Regulate speed of
steam engine
• Reduce effects of
variations in load
(disturbance
rejection)
• Major advance of
industrial revolution
 Ex: CAR Driving
• Consider a car driving
on a curved road
• Under normal driving
conditions the
direction of the car is
continuously adjusted
using feedback from
the driver
 Representing a Control System
• Block diagram shows relationship
between components
• Driver performs numerous functions
including measurement, control and
selecting destinations
Control =
Sensing + Computation + Actuation

GOAL
Stability: system maintains desired operating point (hold steady speed)
Performance: system responds rapidly to changes (accelerate to 6 m/sec)
Robustness: system tolerates perturbations in dynamics (mass, drag, etc)
 Representing a Control System
We can expand the driver block to show
the various functions they perform
 A General Control System
• Many control systems can be
characterized by these components
 Open Loop Control System
An Open Loop control system does not
include feedback Disturbance
 Closed Control System
A Closed Loop control system includes
feedback
A Mechanical Control System
Cistern fill is
controlled
variable
 Float ball acts
as feedback
Valve is closed
as water level
rises
 Mathematical Modeling
• A system model is one or more equations
that describe the relationship between the
system variables – often the input (s) and
output (s) of the system
• For physical systems, these equations are
derived from study of the physical
properties of the system such as
mechanics, fluids, electrical,
thermodynamics, etc.
 Mathematical Modelling
• To understand system
performance, a
mathematical model of
the plant is required
• This will eventually allow
us to design control
systems to achieve a
particular specification
 Mathematical Modelling

The input-output relationship is usually

expressed terms of a differential equation
A Familiar Mechanical Engg. Example

• Spring damper
system
• Mass M is suspended
by a spring and
damper
• Force f is applied
• How do we describe
the motion of this
system y(t)?
Free Body Diagram
Draw an FBD
Insert forces acting on
body
Sum forces
Modern Control System Components

• Process: Physical system, actuation,

sensing
• Feedback: Interconnection between plant
output, controller input
• Controller: Microprocessor plus conversion
hardware (single chip)
 Control System Implementation

Disturbances
Actuator Noise Control I/P Physical Process
Actuators Actual Output
Desired
Control I/P
Sensed O/P
Reference Input Controller Sensors
Sensor Noise
(Desired Output)

Equipped w A/D,
CPU, D/A ports
 Simulating Dynamic Systems
• Dynamic systems are often modeled
using ordinary differential equations or
ODEs
• ODEs are algebraic relationships
between the independent variable and
its derivatives with respect to a single
dependent variable
• MATLAB is a useful computational tool
to simulate dynamic equations
 Simulating Dynamic Systems….
Model: rigid body physics (Ph 1)

Modeling a Spring Mass System

Converting models to state space
form
Construct a vector of the variables that
are required to specify the evolution of
the system
 ROBOT Control System
• How do we determine the motor/actuator inputs so
as to command the end effector in a desired
motion?
• In general, the input voltage/current does not create
instantaneous motion to a desired configuration
– Due to dynamics (inertia, etc)
– Nonlinear effects – Backlash, Friction
– Linear effects - Compliance
• Thus, we need three basic pieces of information:
– Desired joint trajectory
– Description of the system (ODE)
– Measurement of actual trajectory
 ROBOT Control System
desired task (position &
orientation of end-effector)
Actuator
commands
q Actual joint/
.  q end-effector
Trajectory q Manipulator
.. Manipulator location,
planner q control system .
q velocity etc.
.
q q
Control
set-points Feedback
(sensor/observer)

Block diagram of a manipulator control

system
 ROBOT Control System
• Typical single input, single output (SISO) system:

• the robot tracks the desired trajectory and rejects

external disturbances
• the desired trajectory is known, and it is assumed
that the actual trajectories can be measured
• Thus system description is the important for control
engineer
 Operations for a point-to-point motion
control of a manipulator
Tinitial qi
q
Direct and . q
Trajectory q Control 
inverse Manipulator
Tfinal qf generator q.. system .
kinematics q
. .
q q q q

 The interactive effects of inertial,

Coriolis, centrifugal and gravitational
forces vary continuously with time.
 The computation of these from the
dynamic model is complex and time
consuming.
 SISO overview
• We need a convenient i/o description of a SISO
system
• Two typical representations for the plant:
– Transfer function
– State-space
• Transfer functions represent the system dynamics
in terms of the Laplace transform of the ODEs :

• For a 1DOF system :  t  J t  B t 
 s   s 2J s   sB s 
• In the Laplace domain:
 s sJ  B  s 
 s 
P s  
• Transfer function : 1 1/ J
 
 s  s sJ  B  s s  B / J 
 Review of the Laplace transform
• Laplace transform creates algebraic equations
from differential equations
• The Laplace transform is defined as follows:

x s    e st x t dt
0

• Ex: Laplace transform of a derivative:

 dx t  st dx t 

Lx t   L
   e dt
 dt  0 dt
– Integrating by parts:
 dx t 

L   e st
x t  
 s  e st
x t dt
 dt 
0
0

 sx s   x 0
 Review of the Laplace transform
• Laplace transform of a second derivative:
 d 2 x t   st d 2 x t 
Lxt   L 2 
  e 2
dt  s 2
x s   sx 0   
x 0
 dt  0 dt

• For a generic 2nd order system described by

the following ODE:
mxt   bx t   kxt   F t 
• For transfer function representation of the
system:
mLxt   bLx t   kLx t   LF t 
 
m s 2 x s   sx 0  x 0  bsx s   x 0  kx s   F s 
 Review of the Laplace transform

ms 2

 bs  k x s   F s   mx 0  ms  c x 0
• The transient response is the solution of the above
ODE if the forcing function F(t) = 0
• Ignoring the transient response and after
rearrangement
x s  1

F s  ms  bs  k
2

• This is the i/o transfer function and the

denominator is called the characteristic equation
(CF)
 Review of the Laplace transform
• Properties of the Laplace transform
– Takes an ODE to a algebraic equation
– Differentiation in the time domain is
multiplication by s in the Laplace domain
– Integration in the time domain is multiplication
by 1/s in the Laplace domain
– Considers initial conditions
i.e. transient and steady-state response
– The Laplace transform is a linear operator
Review of the Laplace transform
Time domain Laplace domain
xt 

x s   Lx t    e st x t dt
0

x t  sxs   x0

xt  s 2 x s   sx 0  x 0

Ct C / s2
step 1/ s
cost  s / s 2 2

sint   / s2   2
Review of the Laplace transform

Time domain Laplace domain

xt   H t    e s x s 
e at x t  x s  a
x at 
1 s
x 
a a

C t  C
LINEAR CONTROL SCHEMES
• Number of ways by which a controller can react to
the error signal and supply the output for
actuators.
• The on/off or two-step control.
• The proportional (P) control, which produces a
control signal proportional to error.
• The derivative (D) control, which produces a
control signal proportional to the rate of change of
error.
• It is an anticipatory control that measures existing
rate of change of error, anticipates incoming
larger error and applies correction before the
larger error is arrived.
*Derivative control is never used alone.
LINEAR CONTROL SCHEMES
• The integral (I) control, which produces a
control signal that is proportional to integral of
the error with time.
• The integral controller can be considered as
“looking-back” summing all the errors and then
responding.
• Combination modes: Combinations of above
strategies like proportional plus derivative
(PD), proportional plus integral (PI),
proportional plus integral plus derivative (PID)
are often used to achieve the desired
performance.
 LINEAR CONTROL SCHEMES
• The approximate linear model of an individual
joint is obtained by ignoring the configuration
dependent nature of link inertias, interaction
inertias and gravity torques.
• Linear controller based on this approximate
model serves as a good approximation and
results in a satisfactory trajectory tracking
performance for the highly geared (non-direct
drive) joint,
• Such controllers are used in a vast majority of
industrial robots.
• This is basis for the study of more complex
nonlinear control schemes.
 CHARACTERISTICS OF SECOND
ORDER LINEAR SYSTEMS
2
d y dy
a 2 2  a1  a0 y  b0 x
dt dt
a 2 s Y ( s)  a1 sY ( s)  a0Y ( s)  b0 X ( s)
2

The open loop transfer function of this

system is
Y ( s) b0
G( s)  
X ( s) a2 s  a1 s  a0
2

d2y dy Y ( s) b0 n2
 2   2
y  b  2
0 nx G( s)   2
X ( s) s  2 n s   n2
2 n n
dt dt
 CHARACTERISTICS OF SECOND
ORDER LINEAR SYSTEMS
+ E(s) b0 n2
R(s) Y(s)
s 2  2 n s   n2


Y ( s) G( s) b0 n
2
  2
R( s) 1  G( s) s  2 n s  (1  b0 ) n2
The second order linear characteristic equation
of the closed-loop system
 CHARACTERISTICS OF SECOND
ORDER LINEAR SYSTEMS
s  2 n s   n  0
2 2

• The roots of the characteristics equation, which

characterize the system’s time response to a given
input, are
r1 , r2   n  n   1
2

• These roots determine the nature of time response

of the system.
• If roots r1 and r2 are real, then the response of the
system is sluggish and non-oscillatory.
• If r1 and r2 are complex conjugate, then the
response of the system is oscillatory.
 CHARACTERISTICS OF SECOND
ORDER LINEAR SYSTEMS

Legend

y(t)

0.5
= 1
< 1

0
0 2 4 6 8 10 12 14 16
Time, t
 CHARACTERISTICS OF SECOND ORDER
LINEAR SYSTEMS
(i) Damping ratio greater than unity
y(t )  1 
e  n t
2 ( 2  1) 
 
 ( 2  1)   e n ( 2 1) t

 (  1)   e
2
 n ( 2 1) t 


(ii) Damping ratio equal to unity

y (t )  1  e   n t
 n te  n t

(iii) Damping ratio less than unity

e   n t (1   2 
)
y(t )  1  
sin  n (1   ) t  tan
2 1

1 2   
 
LINEAR SECOND ORDER SISO MODEL
OF A MANIPULATOR JOINT...
a a Ba
Motor
1
Link
2 I
Ia

B  

Actuator Reduction Link

gears

Schematic of actuator-gear-link mechanism for a

manipulator joint
 LINEAR SECOND ORDER SISO MODEL
OF A MANIPULATOR JOINT
A common drive system for many industrial
robots consists of an electric motor as the
actuator and a reduction gearbox, to provide
the motion and torque to the joint.

The gear ratio ‘’ is defined as

number of teeth on the gear of actuator shaft η1  L
η  
number of teeth on the gear of link shaft η2  m

η m   L , ηm  L , ηm  L

 LINEAR SECOND ORDER SISO MODEL
OF A MANIPULATOR JOINT...
The Eq of motion can be obtained by
considering gear1, gear2 and link.
For gear1 we have

 m  J mm  Bmm  T1

where,T1 denotes the torque acting on gear1

from gear2 and the link
 LINEAR SECOND ORDER SISO MODEL
OF A MANIPULATOR JOINT...

For gear2 we have

 L  T2  I LL  BLL
where,T2 denotes the torque transmitted to
gear2 by gear1 and τL denote the external
disturbance torque acting on link
From energy conservation
T1 m  T2 L
Since, η is less than one (<1),
 LINEAR SECOND ORDER SISO MODEL
OF A MANIPULATOR JOINT...

 m  TL  J m m   J L m  Bm m   BLm

 2   2

 m  TL   J m   J L  m   Bm   BL  m
2 2

 m  TL  J eff m  Beff m

 LINEAR SECOND ORDER SISO MODEL
OF A MANIPULATOR JOINT...

Since, η is small the effect of the load inertia

and load friction seen from motor is
significantly reduced by factor η2

The effect of load is reduced by η. In a

multilink robot the effect of the coupling
torques due to the motion of other links which
contribute to TL is also reduced.
 LINEAR SECOND ORDER SISO MODEL OF A
MANIPULATOR JOINT...

   M j j   h jk jk  B  G

j j k

  M  B  d


 a   I a   M a   Ba   B a   d
2 2

 LINEAR SECOND ORDER SISO MODEL
OF A MANIPULATOR JOINT...
• Since the configuration
dependent term M is reduced by
a factor , with  <<1
• Ieff is almost constant for highly
geared manipulators.
• This is small compared to direct
drive manipulator (=1)
 LINEAR SECOND ORDER SISO MODEL OF A
MANIPULATOR JOINT...
To summarize the assumptions made
• The links are rigid bodies.
• Link inertia is constant.
• Non-linear interacting terms like the
interaction and centrifugal torques and
gravity torque are ignored.
• Non-linear effects like coulomb friction and
external disturbance are ignored.
• Backlash is ignored.
 LINEAR SECOND ORDER SISO MODEL OF A
MANIPULATOR JOINT...

• Thus, the equation is a SISO model of a

rotary joint of a manipulator as a second
order linear system with actuator torque a
as the input and angular displacement a
as the output.
• To construct a controller for an n-DOF
manipulator, n-independent SISO linear
control systems may be designed.
 JOINT ACTUATORS
Hydraulic actuators,
• produce large force/torque to drive the
manipulator joints without the use of reduction
gearing and are easily applied for robotic
position control.
• cumbersome and messy and require a great
deal of equipment such as pumps, actuators,
hoses and servo valves.
• In applications where position and/or torque
must be accurately controlled, hydraulic
systems prove disadvantageous due to friction
of seals, leakage, viscosity of oil and its
complex temperature dependence.
 JOINT ACTUATORS
Pneumatic actuators
• possess all the disadvantages of
hydraulic systems except that these are
relatively cleaner.
• difficult to control accurately due to high
friction of seals and compressibility of
air (or other pneumatic medium).
• On the other hand, use of hydraulic or
pneumatic actuators to directly drive the
joint minimizes the drawbacks due to
friction, elasticity and backlash.
 JOINT ACTUATORS
Electric actuators
• Although these do not match the power-to-
weight ratio of hydraulic or pneumatic
actuators,
• these are easy to control and interface.
• for small-to-medium sized manipulators.
• require reduction gears of high ratios.
• The high gear ratio linearizes the system
dynamics and reduces the coupling effects.
• increased joint friction, elasticity and
backlash.
 Model of a DC Motor
Ra

+ Armature

ea ia eb .



• the control signal to the DC Motor is

applied at the armature terminals in
form of armature voltage ea.
• The motor torque a is related to the
armature current ia  a  K T ia
 Model of a DC Motor

ea  K b a  ia Ra

Joint Control a Motor Driver ea DC Joint
System (desired) Circuitry Motor 




 Motor dynamics
• When a conductor moves in a magnetic
field, a voltage is generated
– Called back EMF: eb  K 2a
– Where a is the rotor angular velocity
armature inductance armature resistance

dia
L  Ri a  V  eb
dt
 Motor dynamics
• Since this is a permanent magnet
motor, the magnetic flux is constant,
we can write:  m  K1ia  K mia
torque constant
d m
• Similarly: eb  K 2m  K b
dt
back EMF constant

• Km and Kb are numerically equivalent,

thus there is one constant needed to
characterize a motor
 Motor dynamics
• This constant is determined from
torque-speed curves
– torque and displacement are work
conjugates

 0 is the blocked torque

 Motor dynamics
• the ODEs describing this system in
both the electrical and mechanical
domains: L di a  Ri  V  K d m
a b
dt dt
d 2 m d m L
Jm 2
 Bm  Kmia 
dt dt r
• In the Laplace domain:
Ls  R I a s   V s   Kb sm s 
 s 
J s  B s  s   K I s   r
m
2
m m m a
L
 Motor dynamics
These two can be combined to define, the
i/o relationship for the input voltage, load
torque, and output displacement:

integrator
 Motor dynamics
• to express the system as a transfer
function from the input to the output
angular displacement
– But we have two potential inputs: the load
torque and the armature voltage
– First, assume L = 0 and solve for m(s):
Jms 2  Bms m s   I s  Ls  R Jms 2  Bms 
Km
a  m s   V s   K bs m s 
Km

 m s  Km

V s  sLs  R J ms  Bm   K bK m 
 Motor dynamics
• Assuming V(s) = 0 and solve for m(s):
 K bs m s   K m K bs m s   L s 
Ia s    
J m s  Bm s  m s  
2

Ls  R Ls  R r
 m s   Ls  R  / r

 L s  sLs  R J ms  Bm   K bK m 
Note that this is a function of the gear ratio
– The larger the gear ratio, the less effect
external torques have on the angular
displacement
 Motor dynamics
• In this system there are two ‘time
constants’
– Electrical: L/R and Mechanical: Jm/Bm
• the electrical time constant is assumed
to be small compared to the
mechanical time constant
– Thus, ignoring electrical time constant will
lead to :
 m s  Km / R  m s  1/
 
V s  sJ m s  Bm  K bK m / R   L s  sJ m s  Bm  K b K m / R 
 Motor dynamics
• Rewriting these in the time domain gives:
 m s 
J mm t   Bm  K b K m / R m t   K m / R V t 
Km / R

V s  sJ m s  Bm  K bK m / R 

 m s  1/
 J mm t   Bm  K bK m / R m t   1/ R  L t 
 L s  sJ m s  Bm  K b K m / R 

J mm t   Bm  K bK m / R m t   K m / R V t   1/ R  L t 

J B ut  d t 
• By superposition of the solutions of these
two linear 2nd order ODEs:
 Motor dynamics
•the dynamic eq. of a DC motor attached
to a load is written as:
Jt   Bt   ut   d t 
Where u(t) is the input and d(t) is the
disturbance (e.g. the dynamic coupling from
motion of other links)
•To represent this as a transfer function,
take the Laplace transform:

Js 2

 Bs  s   U s   Ds 
 Setpoint controllers
• Let us discuss two initial controllers: PD
and PID
– Both attempt to drive the error (between a
desired trajectory and the actual trajectory)
to zero
• The system can have any dynamics, but
we will concentrate on the previously
derived system
 PD controller
u t   K p et   K d e t 
• Control law:

Where e(t) = d(t) - (t)

• In the Laplace domain: U s   K p  sK d E s 
• This gives the following closed-loop
system:
 PD controller
• This system can be described by:
U s   Ds 
 s   2
Js  Bs
• Where, U(s) is: U s   K p  sKd  d s    s 
• Combining these :
 s  
K p  
 sK d   d s    s   Ds 
• Solving for  gives: Js 2  Bs

Js  Bs  s   K  sK  s   K  sK  s   Ds 
2
p d p d
d

 Js  B  K s  K  s   K  sK  s   Ds 
2
d p p d
d

  s  
K  sK  s   Ds 
p d
d

Js 2  B  K d s  K p
 PD controller
The denominator is the characteristic
polynomial s  B  K  s  K  0
2 d p

J J
The roots of the characteristic polynomial
determine the performance of the system
s 2  2s   2  0
If the closed-loop system is treated as a
damped second order system, this allows
us to choose values of Kp and Kd
• Thus Kp and Kd are:
K p  2J Kd  2J  B
• A natural choice is  = 1 (critically damped)
 Limitations of the PD controller:
– for illustration, let our desired trajectory be a step input and our
disturbance be a constant as well:
 s   , Ds  
C
d D
s s
– Putting this into our system :
 s  
K p  sK d C  D

s Js 2  B  K d s  K p 
– For these conditions, what is the steady-state value of the
displacement?

s K p  sK d C  sD K  sK d C  D K pC  D D
 ss  lim  lim p
 C 
s 0 
s Js 2  B  K d s  K p  s 0 Js 2  B  K d s  K p Kp Kp

– Thus the steady state error is –D/Kp

– Therefore to drive the error to zero in the presence of large
disturbances, we need large gains… so we turn to another
controller
 PID controller
• Control law: u t   K pet   K d e t   K i  et dt

 
U s    K p  K d s  E s 
Ki
• In the Laplace domain:
 s 

 s  
K s
d
2

 K ps  K i  d s   sDs 
Js 3  B  K d s 2  K ps  K i
 PID controller
• The integral term eliminates the steady
state error that can arise from a large
disturbance
• How to determine PID gains
1.Set Ki = 0 and solve for Kp and Kd
2.Determine Ki to eliminate steady state error
• However, we need to be careful of the stability
conditions
B  K d K p
Ki 
J
 PID controller
• Stability
– The closed-loop stability of these systems is
determined by the roots of the characteristic
polynomial
– If all roots (potentially complex) are in the ‘left-half’
plane, our system is stable
• for any bounded input and disturbance
– A description of how the roots of the characteristic
equation change (as a function of controller gains) is
very valuable
• Called the root locus
 PARTITIONED PD CONTROL SCHEME
• A linear controller based on a Partitioned
Proportional Derivative (PPD) control strategy
is slightly different from a simple PD controller.
• It is useful for systems that are more complex.
• The controller is partitioned into two parts: a
model-based portion and a servo portion, such
that the joint-parameters (Ieff and Beff in this
case) appear only in the model-based portion.
• The approximate model of the manipulator
joint developed in previous section is used to
implement this controller.
 PARTITIONED PD CONTROL
SCHEME
Servo-based Model based
Joint
+ (s)
a + (s)
a 1 (s)
a
2
s Ieff
(Ieff s 2 +B eff s)
+ +
Beff s
K d s+K p

E(s)
+ 
(s)
d
 PARTITIONED PD CONTROL
SCHEME
 a  I eff a  Beff  a
 a (s)  ( I eff s  Beff s) a (s)
2

 a ( s)  I eff  a ( s)  Beff s a ( s)   ( s )  s 2  ( s )
a a

•Model based law effectively reduces the

system to that of unit inertia system
•The servo portion of the controller is
based on the error in actuator position
and its derivative.
 PARTITIONED PD CONTROL
SCHEME
a
s
d
2 + a Unit-inertia
system .
 a
Kp Kd

 a (s)  s  d (s)  K d sE (s)  K p E(s)

 2

s  a (s)  s  d (s)  K d sE (s)  K p E(s)

2 2

s E(s)  K d sE (s)  K p E(s)  0

2
 PARTITIONED PD CONTROL
SCHEME
• the natural frequency of the system

n  K p
Kd
• damping ratio  
2 Kp
• the controller for the manipulator should
be critically damped, that is  = 1.

K d  2 K p  2 n
 PARTITIONED PD CONTROL
SCHEME
• Time taken to compute the servo error, the PD
law control gains and to command a new
value of torque, is non-zero and is known as
the cycle time.
• Resulting command torque is a ‘discrete
staircase’ function and the servo error is non-
zero at the beginning of each cycle.
• The controller will reduce this nonzero servo
error to zero during each cycle.
• Hence, the actual trajectory tracked will be
close to, but not exactly the same as the
desired trajectory.
 PARTITIONED PD CONTROL
SCHEME
• Apart from a damping ratio of unity, another
factor that constrains the selection of control
gains is the flexibility of links, which are
assumed to be rigid bodies in the development
of the joint model.
• The unmodeled structural flexibility of the link
and other mechanical elements produces
resonance at frequencies other than natural
frequency, n.
• Since these structural flexibilities have been
ignored, the controller must be designed so as
not to excite these unmodeled resonances.
 PARTITIONED PD CONTROL
SCHEME
Disturbance rejection
• The purpose of control system is to reject
disturbance. If additional input of a
disturbance force fdist.
s 2 E(s)  Kd sE (s)  K p E(s)  f dist
• Let us consider the simplest kind of
disturbance, namely when fdist is a constant.
• Steady state analysis can be performed by
analyzing the system at rest. K e  f
p dist
 Unmodeled flexibility
• Gearing, shafts, bearings and driven links
are not flexible.
• In reality all of these elements have finite
stiffness and their flexibility if modelled
would increase the order of the system.
• If the system is sufficiently stiff the natural
frequencies of these unmodeled resonances
are very high and can be neglected
compared to the influence of the dominant
second order poles that we have modeled
 Unmodeled flexibility
• We must be careful not to excite these
resonances.
• As a rule of thumb, we must limit our closed
loop natural frequencies

 n  0.5 res
This provides us some guideline to choose
gains in controller.
 PARTITIONED PD CONTROL
SCHEME
I0 ω0 is the structural
 res   0 frequency when effective
I max
inertia is I0
In order to prevent exciting these structural
oscillations and also to ensure structural
stability, the controller natural frequency n
must be limited to 0.5 res.

I0
 n  0.5 0
I max
 PARTITIONED PD CONTROL
SCHEME
The control gains Kp and Kd for the PD
controller are selected using following
equation
2 I0
K p  (0.5 0 )
I max

Kd  2 K p
 Effect of External Disturbance
• The performance of the partitioned PD
controller in the presence of an
external disturbance or noise is
examined now
Servo-based Model based
(s)
dist Joint
+
+ (s)
'
a + (s)
a 1 (s)
a
s2 I eff
(Ieff s 2+B eff s)
+ +

Beff s
Kd s+K p

E(s)
(s)
d + 
 Effect of External Disturbance
• The error dynamics in the Laplace
domain
s E(s)  K d sE (s)  K p E(s)   dist (s)
2

• for all bounded disturbances, the

controller will keep the system stable
K K
 dist ( s )  (s  K d s  K p ) E (s) 
2

s s

K
E ( s) 
s( s  K d s  K p )
2
 Effect of External Disturbance
• The steady-state error is defined as
ess  Lim s E ( s )
s 0

• the steady state error in actuator shaft position

K
ess 
Kp
• the steady-state error is inversely proportional
to the proportional gain Kp of the PD controller.
• controller is to provide for disturbance
rejection, that is, to maintain good performance,
Kp is chosen at the maximum possible value
 PID CONTROL SCHEME
• In the PD control scheme the steady
state error caused by a disturbance can
be minimized but not eliminated.
• The control law is modified by adding an
integral term to eliminate the steady-
state error.
• The design of a partitioned PID (PPID)
controller, which is capable of
eliminating the steady-state error
caused by a constant disturbance
 PID CONTROL SCHEME
 a ( s)  I eff  a ( s)  Beff s a ( s)
K i a ( s)
 a (s)  s  d (s)  K d s a (s)  K p a (s) 
2

s
the error dynamics for the PPID controller
is written as
Ki
s E (s)  K d sE (s)  K p E ( s) 
2
E ( s)  0
s
 PID CONTROL SCHEME
Servo-based Model based
(s)
dist Joint
+
+ (s)
'
a
+ (s)
a 1 (s)
a
s2 Ieff
I eff s 2+B eff s
+ +

Beff s
Kd s+K p +Ki /s

E(s)
(s)
d + 

In the presence of a constant disturbance

as a step function of constant magnitude,
dist(s) = K/s
 PID CONTROL SCHEME
Ki K
s E ( s)  K d sE ( s)  K p E ( s) 
2
E ( s) 
s s
K
E (s)  3
s  Kd s 2  K p s  Ki

The steady-state error for the PID

controller in the presence of a
disturbance
sK
ess  Lim s E ( s)  Lim 3
s 0 s 0 s  K s 2  K s  K
d p i
 PID CONTROL SCHEME
• With this control law the system
becomes a third order system
• Often Ki is kept quite small so that the
third order system is close to the
second order system without this term.
Continous vs Discrete time control
• Assumption is - computation of the
control law in zero time (Infinitely fast)
• In place the force is discrete staircase
function
• This approximation is good if the rate
at which new values of f are computed
is much faster than the natural
frequency of the system being
controlled.
• We modeled as n independent second
order differential equations and based
on our controller on that model.
• Computed torque control method
• One method of stability analysis of
nonlinear system known as Lyapunov’s
method.
OTHER CONTROL STRTEGIES
• COMPUTED TORQUE CONTROL
• FORCE CONTROL OF ROBOTIC
MANIPULATORS
• IMPEDANCE FORCE/TORQUE
CONTROL
• Other NONLINEAR CONTROL
Strategies