Warthog Robotics Team Description Paper 2012

Felipe Taha SantAna, Nicholas Makita Fujimoto, Ana Paula Barbosa , Murilo Portela Ribeiro , Yago de Matos Dorea, Douglas Alencar
Av. Trabalhador Sancarlense, 400, 13566-590, So Carlos University of So Paulo, School of Engineering of So Carlos, Department of Electrical Engineering So Carlos, So Paulo, Brazil { fts62019@gmail.com , fujimoto@hotmail.com , menina.ana.barbosa@gmail.com , murilo_portela@hotmail.com , yago.dorea@gmail.com , dssa16@gmail.com }
1

Abstract This paper describes the Warthog Robotics 2D soccer simulation team. The team was runner-up at the Brazilian Robotics Competition (CBR) 2011 and, previously under the name of GEARSIM, won the CBR 2009 and the Latin American Robotics Competition (LARC) 2010. In this paper is presented the current team research focus, which has been designed and implemented within this last year.
I. INTRODUCTION

Ability to approximate non-linear complex system. III. FUZZY SYSTEM FORMATION CHOOSER

The Warthog Robotics 2D soccer simulation team is a branch of the Warthog Robotics Team [1], which was created in 2011, after the merge of two previous robotics groups, GEAR and USPDroids, both competitors at the Brazilian Robotics Competition during several years. Besides the 2D simulation league, the team also participates in two other categories of the Robocup, the 3D simulation league and the SSL. Our approach to the simulation 2D league is the research of decision making in dynamic multi-agent systems. Based on the uncertainty and the subjective character of each simulation 2D game, and on [2], [3], [4], [6], [7], we have chosen a fuzzy system technique to determine the team formation, merging both previous works [6], [7], so that both behavior and positioning of each agents were affected. This method will be presented further on this paper. The Warthog 2D Team is based on Agent2D base source code [5], due its nice and clear implementation of the lowlevel layer, and its easiness in developing intelligent approaches in the high-level layer. II. FUZZY LOGIC The fuzzy systems, introduced by Zadeh in 1973, consist of approaching the human and machine decisions. This makes the machine not only takes exact decisions, like yes or no, but also take intermediate decision: perhaps. The main characteristics of a fuzzy system are the following:

In our previous works, fuzzy systems were implemented on the coach, and they could alter either the positioning of the agents [7], or its behavior [6]. In this new approach, the coach uses both fuzzy systems, blends them and generate a new formation which affect both positioning and behavior, changing completely the attitude of the agents, creating a whole new dynamics to the game. This new fuzzy system has four inputs and two outputs. These parameters are, respectively: - Time - Number of game cycles (Figure 1); - Successful Attacks - Value in percentage (%) (Figure 2); 3); - Successful Defenses - Value in percentage (%) (Figure

- Ball position Value of the X coordinate, where the value 0 is the minimum value, and 105 the maximum value (Figure 4); - Stress Level Represents the behavior to be adopted (Figure 5); Tactical Formation Represents the position to be adopted (Figure 6).

They express inaccuracios and uncertainties; They are systems based on linguistic rules; The reasoning is performed by approximate methods; Conclusions are obtained in parallel; Fig. 1 Total game time [6].

Fig. 2 % Successful Attacks [6].

Due to the mathematical complexity of calculating this new fuzzy system and the limitation of time to perform calculations between each game cycle, we have decided to calculate each fuzzy system separately, and the merge the two results. Comparing with the previous works, instead of having only three behaviors with one positioning, or three different positioning with only one behavior, there are nine possible combinations of formations, as represent on table 1.

Behavior\Formation Stressed Normal Fig. 3 % Successful Defenses [6]. Light

4-4-2 442_S 442_N 442_L

4-2-3-1 4231_S 4231_N 4231_L

4-3-3 433_S 433_N 433_L

Table 1 Behavior and formation It is interesting to point out that both fuzzy systems shares one input (time), and they are influenced by each other. When the formation chosen is the offensive (433), the team tends to attack more, make more shots at goal. This affects the percentage of successful attacks, which affects the choice of the behavior. With more attacks, the behavior chosen tends to be light, which prioritizes a more prudent behavior, with less wrong passes. This generates a more defensive behavior, which causes the ball position (x coordinate) to be smaller, influencing the choice of the formation. IV. TEAM BEHAVIOR A. STRESSED Situation in which the player behaves more aggressively in relation to the opponent, committing more fouls. The stressed behavior is seen in Figure 7, in which the number three player committed a foul, resulting in the punishment with a yellow card. The only goal of a stressed team is goal scoring, If you are in possession of the ball, passes are made rapids, and several long dribbling, ever advancing field. If the opponent is in possession of the ball, marking presses the opponents, too many fouls being committed. There is no concern with any energy spent, because all actions are performed with maximum intensity. With this behavior you get a method of prevention to avoid at all costs the advance of the adversary. B. NORMAL Fig. 6 Tactical Formations [7]. *some parameters are not in English: Muito perto stands for Too close; Perto stands for close; Longe stands for Far away. Situation in which the player has a normal behavior during the match (Figure 8). The team performs this behavior with his actions balancing two conditions: shorter execution action with the greatest energy saving. This form makes the team faster than a defaulting team relaxed, however, will not spend as much energy as a team stressed. His attitude already are bolder, with attempted passes in depth, and a marking tighter. This type of behavior raises some individual plays.

Fig. 4 Ball Position [7]*.

Fig. 5 Stress Level [6].

C. LIGHT Situation in which the player behaves more cautious, avoiding any kind of foul. A team relaxed, as shown in Figure 9, performing the movements with the minimum intensity necessary for successful action. The players move very slowly, dont make faults and their attitudes are always objective and reliable. Passes are made to the teammate that is no opponent around, not doing any solo run, and they run in the balls direction with minimum intensity necessary to arrive before the adversary. This behavior can be seen as a bonus to the team.

V. RESULTS The results were obtained playing ten matches performed as follows: Ten games against the 2010 world champion HELIOS Akiyama (2010b); Ten games against the 2009 world champion WrightEagle Chen (2010); Ten games against the 2010 latin-american champion GEARSIM GEAR (2010). WARTHOGSIM 1 3 3 4 5 6 7 8 9 10 Fig. 7: Stressed behavior AVERAGE 3 3 1 0 0 0 0 1 1 2 0.93 WARTHOGSIM 1 2 3 4 5 6 7 8 9 10 AVERAGE 0 4 1 0 2 3 0 3 3 2 1.17 HELIOS 2 5 4 2 3 3 2 4 2 5 2.80 WrightEagle 1 3 1 0 1 2 2 1 1 1 1.80

Fig. 8: Normal behavior WARTHOGSIM 1 2 3 4 5 6 7 8 9 Fig. 9: Light behavior 10 AVERAGE 0 1 1 2 4 1 3 3 4 2 2.10 WrightEagle 0 1 3 1 2 2 1 1 4 4 1.83

VI. CONCLUSION This paper briefly describes the Warthog Robotics 2D Simulation Team, its current efforts and research areas. We have merged two previous researches in a new fuzzy system that affects both behavior and positioning of the agents. Tests and improvements are still being performed, since the amount of coding needed to implement all the new behaviors were considerably high, but the preliminary results are quite satisfactory. Our plans for the future includes some reinforcement learning techniques applied to the goalie, and a swarm approach for the agents navigation. ACKNOWLEDGMENT We are grateful to all the project supporters, that includes all students who worked with us, our sponsors, the Engineering School of Sao Carlos, and the Institute of Mathematics and Computer Science. REFERENCES
[1] [2] [3] Ross, T.J.:Fuzzy Logic with Engineering Applications, John Wiley, (2004). Nakashima, T. and Takatani, M. and Udo, M. and Ishibuchi, H.:An Evolutionary Apparoach for Strategy Learning in RoboCup Soccer, vol.2,pp.2023-2028, (2004). Nakashima, T. and Takatani, M. and Manabu, N. and Ishibuchi, H.:The Effect of Using Match History on the Evolutionary of RoboCup Soccer Team Strategies, Synmposium on Computational Intelligence and Games, (2006).

[4] Hidehisa Akiyama, Agent2D Simulation League Team, online, available at: http://rctools.sourceforge.jp/pukiwiki/, consulted on January 2010.
[5] [6]

Fraccaroli, ES and Carlson, PM, A Fuzzy Approach For Modeling The Team WarthogSim, (2011). Fraccaroli, ES and Carlson, PM, Time GEARSIM 2010 Da Categoria Robocup Simulation 2D, (2010).