Research On Path Planning And Passing Strategy Of Soccer Robot In RoboCup3D

RoboCup3D simulation match system is a biped robot movement and multi-agent systems (MAS) research and educational tool. Through this platform, we can test the robot theory and algorithms. RoboCup3D simulation match was the new simulation robot soccer competition system introduced in2005. Each player used the Nao model which is bipedal humanoid robot closed to the reality world soccer game player in the year of2007. The humanoid bipedal players are must use feet instead of wheels to walk and run. In this way, compared with the soccer robot with two-dimensional plane, due to the different form of structure, bipedal robot soccer research is facing many new problems and challenges in Robocup3D platform.Biped soccer robots, compared with the two-dimensional planar robot kinematics analysis and path planning have significantly different characteristics. In this paper, the main work as followed:1. Kinematic analysis of simulation biped soccer robotThe kinematic analysis is the base of basic walking and the upper decisions on RoboCup3D simulation soccer game. First through the analysis of simulation the physical parameters of the simulation biped soccer robot, abstract simplify the right leg joint of the simulation biped soccer robot, establish the joint local coordinate system. Then analysis and solution the forward kinematics of simulation biped soccer robot, the establishment of the local coordinate system of each joint to derive the spatial position and orientation matrix derived the trunk location to the end of the hip joint position and orientation transformation matrix by matrix transformation, which determined the position and posture of the end of the ankle, record its position and orientation data. After analysis the structural characteristics of the simulation biped soccer robot legs, using geometric analysis for solving inverse kinematics analysis of simulation the biped soccer robot right leg joints.2. The path planning of simulation biped soccer robotIn the RoboCup3D simulation game system, the world environment is three-dimensional. Traditional plane environment path planning method is no good taking into account the negative impact of the change of movement direction on their balance in three-dimensional space of simulation biped the soccer robot, stability factors that take into account the simulation biped soccer robot high-speed variable to the movement in three-dimensional space, this paper presents a simulation biped the soccer robot path planning method based on genetic algorithm, using reasonable evaluation function, increased to ensure a smooth path to the factors to ensure that the simulation biped soccer robot movement stability, efficiency and smoothness. Group selection methods, improved genetic algorithms and simulated annealing algorithm to ensure the genetic algorithm early species diversity differences and post-competition fierce.3. The passing strategy of simulation biped soccer robotThe passing strategy take a great influence on the competition results in the RoboCup3D simulation game. In this paper, by the application the C4.5decision tree algorithm in simulation biped soccer robot passing training, the players according to the stadium in all aspects of complex factors, comprehensive judgment, pass the ball passing success rate of our players. In the RoboCup3D simulation game system simulation results prove the pass algorithm significantly improve simulation biped soccer robot passing ability.