Soccer robot GGRRT path planning method

Abstract

The invention discloses a soccer robot GGRRT path planning method, aiming to improve the search efficiency of soccer robots and allow the robots to react quickly and take the initiative in the game. Agoal orientation function is introduced on the basis of an original RRT algorithm, so that a growth function is no longer determined by a stochastic growth function alone, but is determined by the stochastic growth function and the goal orientation function together, which is equivalent to increasing the attractiveness of a target point to a research object, and effectively guiding a random treeto grow toward the target direction. Experimental results show that the route planning carried out with the GGRRT algorithm proposed by the invention consumes only 1 / 12 to 5 / 6 of the time cost by theoriginal RRT algorithm, and redundant branches are also greatly reduced, thereby effectively solving the problem of blind search of the soccer robot and greatly improving the efficiency of route planning.

Description

technical field [0001] The invention belongs to an optimization method for robot path planning, and introduces a goal-guided function on the basis of a Rapidly-exploring Random Trees (RRT) algorithm to form a Goal Guide Rapidly-exploring Random Trees (Goal Guide Rapidly-exploring Random Trees) , referred to as GGRRT) algorithm, this algorithm can effectively avoid the blind search of the robot, greatly reduce the search time, is suitable for dynamic and arbitrary complexity application scenarios, and can play an important role in robot soccer games. Background technique [0002] In recent years, with the rapid development of artificial intelligence, robots have once again received unprecedented attention. The football game represented by NAO robots has become a hot sport in which countries compete for the first place, and has also become an important indicator to measure the level of intelligence. In July 2007, NAO was selected as the standard platform by the RoboCup Organiz...

AI Technical Summary

Problems solved by technology

However, the actual results of the game prove that applying the traditional RRT algorithm to the robot soccer game, although it can also ensure the normal and smooth progress of the game, it does not have the strength to win at all.

What is the reason? Because the real-time, accuracy and efficiency of path planning are the core and key of the soccer robot, and the randomness of RRT sampling determines that the robot search process is blind, and the random tree will expand many nodes that are not related to the final path. Only when the leaf node in the random tree contains the target point or enters the target area at a certain moment, can a path from the initial point to the target point composed of tree nodes be found in the random tree. The purpose of the search in the whole process is The performance is not strong, prolonging the time of path search and planning

Therefore, it is not competitive to apply such a set of algorithms to robot soccer games, and it is necessary to improve the RRT algorithm

Images