Research And Realization Of Motion Control Based On Autonomous Robot Soccer

The design of robot soccer is involved in machinery, electronic, automatic, pattern recognition, machine vision and artificial intelligence. The system of robot soccer must be provided with highly real-time and accurate performance. Vision system is one of the important approaches of acquiring environment information, its performance is critical to improve the function of the whole system. In addition, decision and strategy system is the important part of exporting control commands, its control strategy has immediate effect on the intelligent level of robot behavior. This paper mainly studies and probes into the vision system and the decision and strategy system of robot soccer. The main research contents are as follows:1. Vision system.To meet the requirement for obtaining positional information of objects quickly and precisely,a vision system based on omni-vision and front-vision is designed. As for the vision recognition, method of double threshold value based on HLS color space is applied into image segmentation to achieve the goal of recognizing the blob. To correct the distorted image from the omni-vision sensor, the method of transforming image reference frame and extracting arc tangent,means of separately proportion are propounded to get the accurate imformation of angle and dist. For the front-vision, pinhole imaging model of camera is adopted to achieve image localization. And then for getting more precise localization result, the excellent mode of localization is chosen according to the area of blob. The experiment results show the rationality of the vision system and the validity of the proposed localization method.2. Decision and strategy system.The theory of traditional artificial potential field is introduced. The potential field function with adjustable parameters is designed. And this paper analyzes dynamic artificial potential field, which introduces the speed potential force into the potential field function. In this way, the function is able to reflect that the speed of robot, goal and obstacle has effects on the potential field force. The experiment results show that theory of dynamic artificial potential field has great advantages in the aspect of speed adjustment when tracking a dynamic goal. Additionally,the fast method of obstacle avoidance based on omni-vision is propounded. It sets the virtual goal according to the position of obstacle and goal. The robot firstly moves to the virtual goal and then tracks the goal ball. With this method, the high efficient obstacle-avoidance and goal-navigation are realized.