Vous êtes sur la page 1sur 34

Presented By:

Teni Gada- 18.


Jaitheradevi- 29.
Vani Ganji- 20.
Sachin Bhandari- 7.
Sumit Debnath- 65.
Many robotics systems are used in various fields and places in these days.
In order to use robotic systems widely, the robots have to have some
intelligence.
Robots have to adapt themselves to their environment.
The controller of the robots should be able to select the most appropriate
action for a particular situation.
The controller decides upon the most appropriate action by using the Soft
Computing Techniques.
Genetic Algorithms are good at taking large, potentially huge search spaces
and navigating them, looking for optimal combinations of things required
to solve a problem.
A Genetic Algorithm(GA) are guided probabilistic search algorithm modeled after the
Darwinian theory of natural evolution, they are applicable to optimization of controllers.
Search for a optimum program which will solve our problem in an most efficient way.
Initial set of Programs which controls my system.
Analyze output from each Initial Program after applying Function onto it.(i.e. Analyzing
how accurately the program is controlling my system)
Apply Selection algorithm to reject the Program which does not satisfy the minimum
survival criteria( accuracy)
Perform crossover of the random selected program from the program set, creating a new
program having good functionality from the two program which is taken into
consideration for crossover phase.
Add the new program into the set of Initial programs and continue the algorithm till a
optimum program for our controller is created.
GA-learning hierarchical fuzzy control architecture is used in Robot Control.
1) The higher level module serves as a fuzzy classifier by determining spatial features
of arm such as
Straight
Oscillatory
Curved.
2) This information is supplied to lower level of hierarchy where it is processed
among other sensory information such as errors in position and velocity for the
purpose of determining a desirable Control input(torque).
3) The Control input will be in turn given to the robot which generates the e(error) and
(overshoot(value greater than expected one).
Where,
E represents the error in angular position.
-represents the overshoot( output exceeding its final, steady-state value).
Therefore, a fitter individual is an individual with a lower overall error in its time
response.
I am optimizing the fuzzy logic controller which
is used to control my Robot.. Using the Genetic
Algorithm which works on the principle the
fittest survives
Several issues should be addressed when designing a
Genetic Algorithm for optimizing fuzzy controllers:
Design of transformation function.
Incorporating initial expert knowledge.
Choice of appropriate function.
Can a Robot learn to solve problems without being
explicitly programmed?
In other words, how can Robot be made to do what is
needed to be done, without being told exactly how to
do it?"
Genetic programming addresses this challenge by
providing a method for automatically creating a
working program for the Robot.
Genetic programming is a model for testing and
selecting the best choice among a set of programs.
An extension of the genetic algorithm.
The difference between GA and GP is, in GP we use
tree to represent program which is used to solve our
problem.
(+ 2 3 (* X 7) (/ Y 5))
+
2 3 * /
X 7 5 Y
Functions
Terminals
Randomly generate a program that takes two arguments and uses basic
arithmetic to return an answer
Function set = {+, -, *, /}
Terminal set = {integers, X, Y}
1.Randomly create an initial population of individual computer programs
composed of the available functions and terminals
2. Perform the following steps iteratively until a termination criterion is
satisfied
a. Execute each program and assign a fitness value to each individual.
b. Create a new population with the following steps:
i. Crossover: Create a new program by recombining two selected
programs at a random crossover point.
ii. Mutation: Create a new program by randomly changing a selected
program.
3. The best sets of individuals are termed as the optimal solution.
+
Step1:
Randomly create an initial population of individual computer programs
composed of the available functions and terminals.
+
2
Step 1:
Randomly create an initial population of individual computer programs
composed of the available functions and terminals.
Randomly Generating Programs
+
2 3
Randomly Generating Programs
Step1:
Randomly create an initial population of individual computer programs
composed of the available functions and terminals.
+
2 3 *
Randomly Generating Programs
Step1:
Randomly create an initial population of individual computer programs
composed of the available functions and terminals.
+
2 3 * /
X 7
Randomly Generating Programs
Step1:
Randomly create an initial population of individual computer programs
composed of the available functions and terminals.
(+ 2 3 (* X 7) (/ Y 5))
+
2 3 * /
X 7 5 Y
Randomly Generating Programs
Step1:
Randomly create an initial population of individual computer programs
composed of the available functions and terminals.
(+ 2 3 (* X 7) (/ Y 5))
+
2 3 * /
X 7 5 Y
Step 2:
Create a new population using Mutation.
(+ 2 3 (* X 7) (/ Y 5))
+
2 3 /
5 Y
*
X 7
First pick a random node
Mutation
Step 2:
Create a new population using Mutation.
(+ 2 3 (+ (* 4 2) 3) (/ Y 5))
+
2 3 + /
* 3 5 Y
Delete the node and its children,
and replace with a randomly
generated program
2 4
Mutation
Step 2:
Create a new population using Mutation.
(+ X (* 3 Y))
+
X *
3 Y
(- (/ 25 X) 7)
-
/ 7
25 X
Crossover
Step 2:
Create a new population using Crossover.
(+ X (* 3 Y))
+
X *
Pick a random node in each
program
3 Y
(- (/ 25 X) 7)
-
/ 7
25 X
Crossover
Step 2:
Create a new population using Crossover.
(+ X (* (/ 25 X) Y))
+
X *
Swap the two nodes
3
Y
(- 3 7)
-
/
7
25 X
Crossover
Step 2:
Create a new population using Crossover.
Step3:
The program produced is tested for fitness.
If the result is satisfying the fitness criteria we select the program
produced as the best way to solve the problem
Else
Iterate the Genetic Programming steps again to produce a new
program.
1.A single robot is placed at a random position and orientation in a rectangular driving
area closed by walls.
2.A colored ball is also placed at a random position.
3.Using its camera, the robot has to detect the ball, drive towards it, and stop close to it.
4.The robot camera is positioned at an angle so the robot can see the wall ahead from
any position in the field.
GOALS
Find ball in environment
Drive towards it
Stop when close to ball
In the loop, grab an image and analyze it as follows:
1)Convert the image from RGB to HSV (Hue, Saturation and Value).
2)Use the histogram ball detection routine (this returns a ball position in the range [0..79]
(left.. Right) or no ball and a ball size in pixels [0..60]).
3)If the ball height is 20 pixels or more, then stop and terminate (the robot is close
enough to the ball)
Otherwise
If no ball is detected or the ball position is less than 20, turn slowly left.
If the ball position is between 20 and 40, drive slowly straight
If the ball position is greater than 40, turn slowly right.
If the ball height is 20 pixels or more,
then stop and terminate (the robot is
close enough to the ball).
Otherwise
If no ball is detected or the ball position is
less than 20, turn slowly left.
If the ball position is between 20 and 40,
drive slowly straight
If the ball position is greater than 40, turn
slowly right.

Vous aimerez peut-être aussi