Research On Robot Soccer Simulation System

The robot soccer system is one of research focuses in artificial intelligence and intelligence control. The robot soccer system has covered all the fields of robot research, including robotics, machine-electricity, multi-intelligence cooperation, decision making and countermeasure, artificial life and sensor data fusion, image processing, etc. The robot soccer simulation system is a form of the robot soccer system. In contrast with entity robot soccer, the robot soccer simulation system does not need high hardware condition, is easy to implement, and is very suitable for decision-making and planning of robot. The main research of this dissertation includes establishing the decision-making model for robot soccer by using multi-agent programming theory, implementing the action of robots, path planning for robots. The main contents and related original achievements are as follows:The decision-making is the core of the robot soccer simulation system. The dissertation analyses the characteristics of the six-step reasoning model and the three-layer decision model. The agent decision model is established by using multi-agent programming theory. The model is composed of one agent that control and correspond the system, and 11 agents that implement system plans. The agent controlling and corresponding can reason from logic according to the court communication, divide the task into sub-plan, and distribute the sub-plan to agents. According to task and plan, the agents can plan path and confirm the velocity of wheel. The advantage of model includes making the plan of MAS by concentratingand agents implementing sub-plan by distributing.The planning path research of robots is always the focus of robotics. At the same time, it is an important part of robot soccer simulation system. The way of planning path includes whole-area planning, sub-area planning and mix-area planning. The way of whole-area planning plans the path according to the communication of the whole-area. The way of sub-area planning plans the path according to the communication of the sub-area. The dissertation researches the way of whole-area planning including genetic algorithm, grid and artificial potential field. With analyzing the advantages and disadvantages of those planning way, whole-area planning is suitable to complex circumstance, but it cannot deal with the real-time and dynamic circumstance. According to the decision system, the way of sub-area planning is implemented by using path planning based on geometry algorithm. Robots pass the obstacles with limit-cycle. Path planning based on geometry algorithm can deal with the real-time and dynamic circumstance. The result of simulation indicated that geometry algorithm is suitable to the sub-area planning path.The behavior of robot is an important part of the robot soccer simulation system, that is executor of system strategy and path planning. The basic actions are implemented by adjusting the velocity of left-right wheels. The basic behaviors are summarized, including shooting by using the way of perpendicular bisector, adjusting direction in the course of moving to the ball, and making the moving ball to be controlled with dichotomy. The basic method of shooting is improved, in which robots shoot to the blind area of goalkeeper, so that the rate of field goal is enhanced.