Design of robot soccer simulator

Robot soccer game is an interesting emerging domain for multiple cooperative robotic system. This paper discusses the detailed design of a simulator, and describes the architecture of soccer server and client in detail. This simulator is sufficiently flexible and robust for the users to develop strategies for a simulated competition and to test algorithms of intelligent robotics.

Robotic velocity generation using neural network

The fast paced nature of robotic soccer necessitates real time sensing coupled with quick decision making and behaving. The robot must have high response rate, exact motion ability, and must robust enough to confront interfere during drastic match. But during the match, we find that the robot usually do not act exactly as the commands from host computer. In this paper, we analyze the reason and present a method that uses BP neural network to output robotic velocity directly instead of conventional path plan strategy, to reduce the error between actual motion and ideal plan.

Six-step reasoning model for robot-soccer

The decision making system of robot soccer is a kind of knowledge system. A Six step Reasoning Model is established by formalizing its expert knowledge and decision making process. Furthermore, many other models can be considered as mutation and evolution of the Six step Reasoning Model.

Algorithm on robot soccer pass and shoot a ball

Introduces of robot soccer’s competition software of Harbin Institute of Technology (HIT), the concept of running range and the method of calculating the running range for both the opponent and our teammates according to the distances between the ball and robot soccers, and therefore the method of calculating the angle that the robot passes or shoots the ball according to the running ranges of both sides. And gives the examples of passing the ball when the ball’s position is in the backcourt and shooting the ball when the ball’s position is in the frontcourt.

Behavior selection strategy for soccer robots

The soccer robot system and the effective multi agents cooperation strategy applied to the MASKARO team composed of a set of action controllers, a set of behavior module and a behavior selector are introduced. The action is the primitive low level component of the robot control system necessary to move the robots on the playground. Each action controller determines the linear and angular velocity commands of the robots corresponding to its own purpose. The behavior is the high level component of the robot control system composed of necessary action sequences. Each behavior module determines the desired action sequences and action commands corresponding to its own objective. The behavior selector considering the information that comes from the vision system selects the behavior of each robot every sampling time. Thus, the behavior of each robot is changed dynamically. The presented strategy is successfully applied to the MASKARO team and the team is ranked in the first place in the 2000 FIRA Korea Cup K League.

Robot soccer action selection based on Qlearning

This paper researches robot soccer action selection based on Q learning .The robot learn to activate particular behavior given their current situation and reward signal. We adopt neural network to implementations of Q learning for their generalization properties and limited computer memory requirements.

Development of RoboSot category soccer robots

In the RoboSot category, no central station or over head camera is allowed. Therefore several issues such as sensors, mechanisms and CPU boards must be taken into consideration before building an autonomous and intelligent system for this category. Each robot must have appropriate sensors and must be able to communicate with other robots to have timely information concerning the ball and robot position in the field. Computing power also becomes and improtant factor in implementing intelligence and autonomy onboard. These issues are discussed. Then the robot structure of the Dreams Come True(DCT) from KAIST is introduced. Each robot has a Pentium Π single board computer, and Ethernet card, and USB camera and an omni directional mobile mechanism.

Design of vision-based soccer robot using DSP

A new design of vision based soccer robot using the type TMS320F240 of DSPs for MiroSot series is presented. The DSP used enables cost effective control of DC motor, and features fewer external components, lower system cost and better performances than traditional microcontroller. The hardware architecture of robot is firstly presented in detail, and then the software design is briefly discussed. The control structure of decision making subsystem is illuminated also in this paper. The conclusion and prospect are given at last.

Recognition of robot-soccer in point of information fusion

Summary of typical information fusion systems, synthesis analysis of Robot Soccer’s architecture, recognition of its characteristic and key technique are given. The result is prompted that Robot Soccer can be treated as a platform of the information fusion.

Layered learning of soccer robot based on artificial neural network

Discusses the application of artificial neural network for MIROSOT, introduces a layered model of BP network of soccer robot for learning basic behavior and cooperative behavior, and concludes from experimental results that the model is effective.

A Research Platform of Multi-agent System Robot Soccer Game

A soccer robot system (HIT 1) was built to participate in MIROSOT_China99 held in Harbin Institute of Technology. Robot soccer game is a very complex robot application that incorporates real time vision system, robot control, wireless communication and control of multiple robots. In the paper, we present the design and the hardware architecture and software architecture of our distributed multiple robot system.

Building Integrated Mobile Robots for Soccer Competition

Robot soccer competition provides an excellent opportunity for robotics research. We have built a soccer robot system to participate in internal and oversea matches. Firstly, we propose a new learning control scheme adaptive PID learning controller. It means to overcome the drawbacks of the conventional PID type control methods. Secondly, we introduce our vision recognition algorithm. It remarkably increases the speed of recognition. Finally, we refer the communication system. We adopt bulletin board system to prevent communication confusion.

A global vision system: using hue thresholds to exact feature and recognize

The fast paced nature of robotic soccer necessitates real time sensing coupled with quick behaving and decision making. In the field with real robots, it is important to well perceive the location of ball, team robots and opponent robots through the vision system in real time. In this paper the architecture of global vision system of our small size robotic team and the process of object recognition is described. According to the study on color distribution in different color space and quantitative investigation, a method which uses H (Hue) thresholds as the major thresholds to feature exact and recognize object in real time is presented.

Robot path planning using genetic algorithms

Presents a strategy for soccer robot path planning using genetic algorithms for which, real number coding method is used, to overcome the defects of binary coding method, and the double crossover operation adopted, to avoid the common defect of early convergence and converge faster than the standard genetic algorithms concludes from simulation results that the method is effective for robot path planning.

Decentralized algorithm of Kalman filtering

Presents an algorithm which can be used to achieve complete decentralization of Kalman filter algorithm amongst sensing nodes of a multi sensor system, and points out the algorithm can be used for position estimation in Robot Soccer because it does not require any form of central processing facility or centralized communications medium, and illustrates with a simulation example that it is very effective.