Motion Planning Of Humanoid Soccer Roboit

Humanoid soccer robot as an important branch of robotics research has all the features ofthe humanoid robot, it has beng a focus topic for scholars all over the world. In this thesis, themain research is the humanoid soccer robot path planning problem. In view of the ant colonyalgorithm has parallel search ability, but it shortcomings search early slow convergence, thisthesis finally introduce the genetic algorithm to improve, and applied in solving the problemof path planning.Firstly, the grid method is used to devide the humanoid soccer robot environmentalspace, for special obstacles may appear in the environmental space lead to the ants into deadcirculation using the presents convex processing theory to solve. And using ant punishmentstrategy to solve when turning more paths appear in the path planning.Secondly, this thesis made some improvement on ant colony algorithm, in the analysisof ant colony algorithm to search the reason of slow early in the algorithm, found the causesfor this problem is because early in the algorithm, the pheromone secretion insufficiency, butthe genetic algorithm has the characteristics of fast search in the prophase, the introduction ofgenetic algorithm to determine the information of the ant colony algorithm pre the hormonesecreted, and at last determine the "population evolution reached maximum generation set; orcontinuous generations, combined with the timing of progeny populations evolution rate arelarger than the maximum rate of evolution" as the standard of the two.Finally, using MATLAB to simulate the system, and the simulation results achieve goodexpected effect, fully demonstrates the feasibility of the new method.