The optimal design of Soccer Robot Control System based on the

advertisement
Research on the Optimal Design of Soccer Robot based on the
Mechanical Analysis
Menghai LIANG1, a, Gang XU2, b
1
Department of Electrical Engineering and Automation, Luoyang Institute of Science and
Technology, Luoyang, 471023, China
2
Department of Electrical Engineering and Automation, Luoyang Institute of Science and
Technology, Luoyang, 471023, China
a
email: lylglg@163.com, bemail:kxu20112012@163.com
Keywords: Soccer Robot; Mechanical Analysis; Optimal Design
Abstract. According to the problem that the intelligence is not high of the soccer robot, using the
mechanical theory as a guide, making some mechanical analyses and calculations on the pressure
and transmutation states of chip kick mechanics, and conducting optimal design too, then making
the structure of chip kick mechanics more and more rationalization. Experiments show that the new
soccer robot controller features a quick response and high servo rigidity, and provide a kind of
method for improving and perfecting the soccer robot control system, at the same time, filling the
needs of producing.
Introduction
Robot World Cup [1][2] is a form of football, through the competition to test new technologies
in order to promote artificial intelligence, robotics and development projects of related areas;
meanwhile, it also provides a broad platform to the robot hardware, software technology research.
Robot soccer game is put forward in recent years of multi-agent system development platform, and
it is a typical multi-agent robotic System. Soccer robot design has become the artificial intelligence
and robotics research in the one of the hotspots field [3].
Soccer robot control system [4] [5] as an executive structure control system, the quality of its
performance largely determines the outcome of the game; it directly affects the accuracy and
flexibility of the robot motion, and the reliability of the whole system [6]. Because of the
intelligence of soccer robot be designed at home and abroad is often not high, It can not meet the
needs of the robot control system [7]. Therefore, the development of a high-performance control
system has become an urgent desire [8].
The control system working environment is dynamic, and the other parties of the robots move
are often unpredictable. This requires the robot can not only control their players accurately, but
also can actively control and track the ball [9]. The mobile robot design and manufacture
technology, motor drive control, sensors and the requirements of the artificial intelligence are very
high. Many researchers have been committed to perfect and improve the performance of the robot
[10].
In this paper a new optimal design of soccer robot control system which is based on mechanical
analyses and calculations on the pressure and transmutation states of chip kick mechanics, this new
control system with high precision for speed control and high dynamic quality.
Design of the Chip Kick Mechanics
At present the main means is low click type and the lever type, low click type is on the bottom of
the ball through attack the ball flew over obstacles, this method is able to pick the ball’s advantages
and makes the energy loss in institutions least, the shortcoming is the ball high requirement of the
shape of the electromagnetic valve [11]. Therefore, the development of a high-performance control
system of soccer robot has become an urgent desire for soccer robot fans.
With the continuous development of robot technology, the team at home and abroad with the
basic function can play football and in gradually to pick the ball function development. Pick the ball
design break through the traditional 2 d game model, this makes robot can directly hit the opponent
ball and don't have to go around, this simplified decision-making, increase attack and breakthrough
are quite useful. But because of the limits of the robots, how to make robots in the limited space in
both playing and picking the ball is a big test of the structure design.
Pick the ball mainly uses in the robot a throw-in, pass the ball and goalkeeper big foot clearance
[12]. According to the rules and regulations, the robot can't longer than 0.016m, and then according
to the movement method, pick the ball’s length and height demand for: Throw in ball. Rules
defenders to leave at kick-off 0.5m, but it has to pick the ball out and shoot above 20°. Pass a
similar to shoot angle of goalkeeper clearance, but distance is not so strict requirements. Because
the others can close to our robot, so pick the ball out shoot the bigger angle the better, and the
farther the distance the better. According to the demand of the game, according to its own tactics
making soccer robots pick the ball institutions put forward the following requirements, pick the ball
maximum height of 0.035m, pick the ball the furthest distance 3 m, pick the ball out of the shot
more than 38° angle.
It is known that the cue ball between pick the ball rod have large enough frictional force, and the
role of the electromagnet only a few milliseconds process, so we can assume that between the ball
and pick the ball device without relative motion, at this time, pick cue ball and angle α turn after
pick cue separation. Pick in the cue ball club, the fulcrum and the ball contact point and ball center
composed of triangle relations can export:
l 2  w2
r
R


sin(   ) sin  sin 
        
(1)

(2)
2
In the type: l is the fulcrum height; w is picking out the length of club; r is the fulcrum which is
to centre distance; R is QiuBanJing. β is Ball shoot out angle. From the laws of rotation we can find:
J   F ( L  l )  mc g (l  h)sin   mb gr cos(      )
(3)
In the type(3): L stands for pulling force point role of height; mc stands for pick cue quality; h
stands for pick cue censored height; mb stands for the ball quality. Request to be tight with the ball
club and the friction force ball etc. By kinetic energy theorem we can find:
1
Wk  mc g (l  h)(1  cos  )  mb gr[sin(    )  sin(     )]  (mc  mb )v 2
(4)
2
(v sin  )2
(5)
H max 
2g
v 2 sin(2 )
(6)
Smax 
g
In the type, Wk stands for electromagnet work, which can be used to get the ball injection speed v,
when the ball with speed v and Angle β from robot, Hmax is flight level, Smax is fall place distance.
Thus, we can get the ball through the calculation of pick out the angle, and pick out the speed, pick
the ball and pick the ball high furthest, etc.
Institutions Optimization Design
The ball R=21.6mm, mb =45g. To ensure the cue ball and pick and not run into low contact
choose other parts of the cue for w=16mm. From the law of conservation of energy, it is known that
the lighter lift stick, the faster the speed at first, consider to pick cue the deformation of stress
intensity factor, finally optimization for mc=34g.Since robot have play ball institutions, in order not
to decorate with the wheel position conflict, only can pick the ball in play on electromagnetic valve
electromagnet institutions. According to the electromagnet diameter and robot’s height, take
L=58mm, h=19mm.
According to the above theory formula, let’s take the part of determined and the other institutions
into the condition, and let the rest of the adjustable parameters for quantitative analysis of the
design. It can be found to pick the ball with the organization of the contact point has a best value.
The point of the height and pick the ball institutions can achieve maximum distance, and the
position of the ball fulcrum makes the lift distance and the lift height maximization produce
different places, so we can't thought set lift ball point to satisfy the high altitude and the maximize
the farthest distance at the same time. Electromagnet for height and the distance of the influence is
linear, so we should as far as possible to improve the average output power electromagnets.
According to the front of the quantitative analysis, set to pick the ball distance as far as the
objective function is as follows:
2
sin(2  )
(7)
max F ( X )  [
(mc g (l  h)(1  cos  )  mb gr (sin(    )  sin(      ))  Wk ]
mc  mb
g
Design variable X= [β, l, H] . Set constraint conditions: shot out angle 45°<β<90°; Pick the ball
institutions fulcrum height 0.049 m<l<0.59m; lift the ball institutions and the ball the contact point
high 0m<H<0.049m.The optimization mathematical model belongs to the constraint
single-objective nonlinear optimization problems. Choose SUMT penalty function method to get
the best value for all variables, this method through the tectonic augmented objective function, and
the constraint optimization problem into a series of unconstrained minimization problems, through a
series of questions to solve constraint and to get original constrained problems optimal solution.
Definition function is:
T
m
2
F ( x, a)  f ( x)  a  [max(0, gi ( x))]
(8)
i 1
The initial plan for x0=[45,0.049,0]T. After 17 iteration calculation, and get the optimal solution:
l =0.043m,H=0.0038m,Smax=4.38 m.
Test results
Put the calculations parameters into Matlab ball movement track simulation, the results shown in
Figure1 (a). Among them, the longest flight distance is 4.312m, top flight is 1.102m, and the flight
time is 0.896s. In physical experiment, with corresponding image processing program acquisition
lift the ball a parabolic height and distance data, through the parabolic curve fitting the obtained
data see Figure 1 (b) and (c)
(a) The theoretical curves
(b) chip kick mechanics
(c) No chip kick mechanics
Fig.1. The experimental results
The ability of pick the ball robot has reached theoretical calculation expected. The experiment
started with no the institutions, the ball can not very well joint with pick cue, causing pick the ball
dynamics change range is very large, and pick the ball height and the average of the theoretical
calculation of the distance is smaller than the calculated assumption the height and distance.
After the ball add tape loading agencies, institutions force the ball the reverse spin, the ball
close to pick the cue, pick the ball transfer fully energy to the ball, pick the ball effect is obvious
stable. But it is found that the actual pick the ball after add tape loading agencies less than the
theory calculated average distance.
Conclusion
Based on the analysis of the mechanical theory as the foundation, designed the soccer robot pick
the ball institutions optimal design process, found aim function, select design variables and the
corresponding optimization algorithm to optimize a complete set of institutions. At last through the
test to get the final performance parameters of the institution. Experiments show that the system has
higher accuracy and stability, the new optimize pick the ball have design basic requirements, and
achieved good ideal control effect.
Acknowledgement
In this paper, the research was sponsored by the Nature Science Foundation of Henan Province
(Project No. 201112400450401) and Youth Fund Project of Luoyang Institute of Science and
Technology (Project No. 2010QZ16).
References
[1] Satoshi Kagami,Tomonobu Kitagawa,Koichi Nishiwaki,Tomomichi Sugihara,Masayuki
Inaba,Hirochika Inoue. A Fast Dynamically Equilibrated Walking Trajectory Generation Method
of Humanoid Robot [J], 2002.
[2] HIROKIK, MINORUA, YASUOK, ET al.RoboCup: a challenge problem for AI and robotics.
HirokiK. RoboCup-97: Robot Soccer World Cup [C]. Berlin: Springer, 1998.38-43.
[3] Zhizhong Yin. Application of FPGA control DC motor servo system [J]. Inner Mongolia
Science and Technology and Economy, 2008 177 (23) 101-103.
[4] Joerg Christian Wolf, Phil Hall, Paul Robinson, Phil Culverhouse. Bioloid based Humanoid
Soccer Robot Design, 2007.
[5]Wu Chuan-yu, He Lei-ying, Design and Realization of Instructional RPPR-Robot, Research and
Exploration in Laboratory.2007, 26(10)
[6] Nishiwaki K,Kagami S. High Frequency Walking Pattern Generation based on Preview Control
of ZMP, IEEE International Conference on Robotics and Automation. 2006.
[7] Chen Nan and so on. For the field of industrial and highly interconnected, TI launched a new
Sitara ARM9 microprocessor. [J]. Global Electronics, 2010 (5) 86-87.
[8] Vision Heading Navigation Based on Navigation Curve [A]. Proceedings 2010 International
Conference on Intelligent Computing and Integrated Systems[C]. 2010.
[9] Chen Nan and so on. For the field of industrial and highly interconnected, TI launched a new
Sitara ARM9 microprocessor. [J]. Global Electronics, 2010 (5) 86-87.
[10] Jonghoon ZMP Park,Youngil Youm.General ZMP Control for Bipedal Walking. IEEE
International Conference on Robotics and Automation. 2007.
[11] Chunmei Xu. Mechanical servo system based on fuzzy neural network for complex control [J].
Control Engineering .2010:17 (2):146-148.
[12] Zhijun He. LM629-based motor servo control system design [J]. Mechanical design and
manufacture .2009 (2) 40-42.
Download