A METHOD TO CREATE THE MOST ACCURATE TARGETING

IN 1 ON 1 SOCCER ROBOTS

Yahya Hassan Zadeh Nazarabadi, Hossein Saghlatoon and Mohammad Ali Sharif Shazileh

Parse United Robotics Team, Ferdowsi University, Park Sq, Mashhad, Iran

Keywords: C Programming Language, 1 to 1 Soccer robots, Ultrasound, Compass, Microcontroller.

Abstract: Because the final state of robo soccer is targeting the goal, it’s very important to design and use standard

high accuracy methods. In this paper we discuss about goal targeting and design a method for robo soccer to

target more accurate than the other methods. This method is easier, less complicated and needs less space on

processor flash memory.

1 INTRODUCTION

The main goal of robocup is to, by the year 2050,

develop a team of fully autonomous humanoid

robots that can win against the human world soccer

champion. Up to now, many fields of soccer robot

have been developed such as Humanoid Robots ,

Small & Middel Size Robots , 1 on 1 and 2 by 2

soccer Robots.

This is why soccer robots must be autonamous .

Designing their algorithm is disscused as the most

important aspect.

Therfore a research has been done to design an

optimum and more accurate with the least slip

algorithm for goal targeting in 1 on 1 soccer robots.

In this event, 1 on 1 autonomous mobile robot

team plays in a highly dynamic environment,

tracking a special light-emitting ball in an enclosed,

landmarked field.

In this paper we first define the problem in part

2, in part 3 we have a discussion about method and

materials, in part 4 we introduce the results of

testing this method with a robot , and finally in part

5 we come to a conclusion and give suggestion how

to develop this alghorithm to other kinds of soccer

robot.

2 DEFINING THE PROBLEM

In this section we speak about the problem and use

the global standard rules of 1 on 1soccer robots

made by Robocup organization and used in

international robocups.

Two different kind of fields, named SOCCER A

and SOCCER B, may be used at a tournament.

SOCCER A: The playing-field is 122 cm by 183

cm. The corners are flattened.

SOCCER B: The playing-field is 122 cm by 183

cm. There is an open space of 30 cm width around

the field.

Walls are placed all around the field, including

behind the goals and, if applicable, the open space.

The height of the walls is 14 cm. The walls are

painted matte black.

SOCCER A: The width of each goal is 45 cm,

centered on each of the shorter sides of the playing-

field. The goal is 14 cm high. It has a cross-bar on

top (to prevent robots from entering the goal). The

interior of the goal including floor, walls and cross-

bar painted, one side yellow and the other side blue.

The exterior is painted black.

SOCCER B: The width of each goal is 60 cm,

centerd on each shorter sides of the playing-field.

The goal is 10 cm high. It has a cross-bar on top (to

prevent robots from entering the goal). The interior

of the goal including floor, walls and cross-bar

painted, one side yellow and the other side blue. The

exterior painted black.

The fields should be placed such that the

influence of external infrared light be as low as

possible and the magnetic field of the earth is

disturbed as little as possible. Perfect conditions

cannot be guaranteed, however.

269

Hassan Zadeh Nazarabadi Y., Saghlatoon H. and Ali Sharif Shazileh M. (2010).

A METHOD TO CREATE THE MOST ACCURATE TARGETING IN 1 ON 1 SOCCER ROBOTS.

In Proceedings of the Fifth International Conference on Evaluation of Novel Approaches to Software Engineering, pages 269-272

DOI: 10.5220/0002932102690272

 SciTePress

Figure1: A standard Robocup field.

A well-balanced electronic ball shall be used. The

ball will emit infrared (IR) light.

Figure 2: The standard ball with IR Sensors.

So robot should have IR sensors in different vectors

to detect the ball and come to get it.

But the main problem discussed in this paper is

after detecting ball and getting it, How we can detect

the goal and shoot to the goal with high accuracy

and make sure that our robot is in a suitable

situation or not?

3 METHODS AND MATERIALS

In this section we discuss the algorithm we use to

have an accurate goal detecting and shooting.

3.1 Situation Detecting Hardware

First we should define the materials use to detect our

situation. In this research a compilation of spherical

& cubic coordinate is used to locate the situation. So

distance detecting and angle detecting sensors are

needed.

We use ultrasound sensors for distance and apply

a compass sensor to detect angle.

Notice that because of special kind of ball any

IR-based sensor that is used for distance or angle

detection is not accurate enough.

Figure 3: Ultrasound Sensors.

Figure 4: Compass Sensor.

We also need a processor to process the output

signal of these sensors.

Figure 5 shows a soccer robot made by the

author of this paper. This robot has the same

hardware and algorithm.

Figure 5: Soccer robot.

3.2 Situation Detecting Algorithm

To detect the situation we should use our output

signals of sensors and process them.

As default we use 4 sensors in robot, one in each

side of robot (east, west, north and south). Also the

zero degree of angle is shown in figure 6.

In this method robot finds the target with its distance

to the walls of field and its angle. Indisputable for an

accurate shoot, robot should turn to appropriate

angle ratio to goal .since the angle is variable into

the situation, we use the total distance to walls to

find appropriate angle. In this way robot first finds

its situation and then finds the accurate angle for

shooting the ball (shooting angle), then the robot

rotates to the proper angle and shoots the ball.

ENASE 2010 - International Conference on Evaluation of Novel Approaches to Software Engineering

270

Figure 6: Robot’s Ultra sensor and 0 degree.

3.2.1 Single Situation Detection

In this way , first suppose that just one robot is in the

field, the total distance of robot to the walls in x & y

are constant extent and equal to width and length of

the field. Another alternative is when we calibrating

the robot in the field it defines the front gate as

rival's and the behind as it's own. So according to

figure 6, the angle in attack normally varies between

- 90 degree (right) to +90 degree (left).

To define coordination as shown in figure 7,

robot has 4 numbers to show its coordination.

Variables x1 and x2 are to define coordination in x

vector and variables y1 and y2 are to define

coordination in y vector.

Figure 7: coordination parameters.

The black box is the target and the violet line is

drawn from the center of the target. Another line is

drawn from the center of target to the center of

robot’s shooter device. Angle A is the appropriate

angle for shooting. In other words, if robot rotates

from its 0 to A and then shoots the ball, it will hit the

target.

To calculate the amount of A in this method

robot first should denote whether it's in the

right-half of field or left one. Here is this detection

code in C++ programming language.

Int cor ;

Float x1,x2,B,angle;

if(x1>x2) cor=1;//denote right

else if (x1<x2) cor=2;//denote left

else cor=0;//denote middle

switch (cor) {

case 1:

B=L/2-x1;

Break;

Case 2:

B=L/2-x2;

Break;

Case 0:

B=L/2;

};

After defining the robot area, now it’s the time to

calculate angle A. As shown in figure 7 we suppose

that X2 > X1 so according to upper code cor = 2.

Other supposition are the Length (=L) and

Width (=W) of the field. So as shown in equations

bellow:

B = x1 – (L/2) (1)

A=arctan(y1/B) (2)

According to Figure 6 if x1 > x2 we should

subtracting A to -1 and if x1 < x2 we should subtract

A to 1.

So according to equations (1) and (2) robot can

calculates the amount of angle A.

In other words this method consist of 5 stage:

1) Robot should detect the ball and go to get it.

2) after getting the ball, Robot should turn to its 0

degree.

3) after turning to zero, robot should calculate the

appropriate angle (A) .

4) after calculating the angle, robot should turn to

that angle .

5) shooting and Goal!

A METHOD TO CREATE THE MOST ACCURATE TARGETING IN 1 ON 1 SOCCER ROBOTS

271

3.2.3 Situation Detection in Match

Because in 1 on 1 soccer there are only two robots in

the field, this method suffers some error states that

cause problems for robot .These stages happen when

one robot is sensing by ultrasound sensors at the

time that the other robot wants to shoot. In this

occasion the stricker robot should move to change

its situation. This scape from the last occasion

depends on the robot’s velocity and the clock of

system.

3.2.3.1 How to Detect this Error Situations

As shown in figure 7 we have equations below:

x1+ x2 = L (3)

y1+ y2= W (4)

In fact according to the situation two variables of x1,

x2, y1, y2 are as data and the others are as parity. In

other words, before using this method to notice that

whether robot is in the error state or not, it should

use the equations 3, 4 as provision to continue the

method.

4 RESULTS

This method has been tested with the soccer robot

shown in figure 5. The field was

182 (=L) * 122 (=W). The coordination, targeting

angle and average of accuracy in 100 times of trying

out (20 times for each one) shown in table 1.

Table 1: The results of testing this method with a robot.

Number X1 Y1 Angle Accuracy

1 31.6 59.3 0.772284 100%

2 42.1 96.7 0.601476 100%

3 132.9 100.2 -1.18239 100%

4 91.6 90.2 -1.56969 100%

5 60.0 88.2 1.227777 100%

5 CONCLUSIONS

In this paper we first mooted a problem in targeting

goal in 1 on 1 soccer autonomous robots. Then we

designed a method to have a most accurate targeting.

In this method robot in each situation could

easily finds the appropriate targeting angle to the

center of goal.

By using this method robot has a most accurate

targeting than the other methods. As shown in table -

1 the accuracy of this method for 100 times of trying

out is 100%.

Though this method is designed for 1 on 1 soccer

robots, but it could be used in other soccer robots

like small size or middle size leagues.

REFERENCES

Hassan Zadeh Nazarabadi, Y., Yazdanpanah, N, 2009.

Design the algorithm of deminer autonamos robot. 1

National Conferance On Fundamentals Of Software

Engineering, Lahijan, IRAN.

Floyd, 2006. Digital Fundamentals, Pearson Education.

London, 9

edition.

Kaheh Ali, 2009. AVR microcontroler and their

fundamental, NAS, Tehran, 4

edition.

ENASE 2010 - International Conference on Evaluation of Novel Approaches to Software Engineering

272