Path Planning Of Soccer Robot And Its Countermeasure

Soccer robot system is a large-scale system integrating environment perception,dynamic decision,behavior control and behavior execution,it is a combination of robotics and artificial intelligence.Path planning with confrontation decision-making is an important part of soccer robot decision-making system.The path planning aims at finding the optimal path,and completes the dynamic and static obstacle avoidance.The decision algorithm can be used to make the strategy of soccer robot and the choice of cooperative strategy quickly.This topic mainly takes the soccer robot competition as the supporting scene,studies the path planning by using improved artificial potential field method and particle swarm intelligence algorithm,studies the strategy of robot soccer robot by using the comprehensive strategy of rough set theory.First of all,when the obstacle is near the target point,using the traditional artificial potential field method in the path planning,there are some problems of that the soccer robot can not reach the target and the local minimum value.Using the basic particle swarm algorithm,the soccer robot can reach the target but the algorithm would appear premature and poor convergence,so the two algorithms are improved.In order to solve the problem of target-unreachable and local minimum in path planning,the potential field function method of relative distance between robot and target are introduced to improve the artificial potential field method by changing the resultant force.The problem of premature convergence and poor convergence,a fitness function and a security constraint function is introduced to improve the algorithm.Solving the obstacle avoidance effect of particle swarm optimization algorithm is not obvious problem,the nonlinear dynamic inertia weight decreasing law is used to improve the algorithm.In the path planning of the improved artificial potential field method,the soccer robot can successfully avoid the obstacle reach the target point.But when the parameter is not set properly,the path is not the optimal.In the path planning of the improved particle swarm algorithm,not only the robot can avoid obstacles to reach the target point smoothly,but also to achieve the shortest path,obstacle avoidance effect is better,but if the learning factor is set badly,the obstacle avoidance effect is poor.After adding the self-adaptation,path planning for improved particle swarm optimization can independently learn to complete the selection of the value of learningfactors.In the face of obstacles in the direction of the moving obstacles and the pursuit of the target point are able to avoid obstacles in real time and the effect is obvious.Experiment simulation and comparison between in improved potential field method and particle swarm optimization algorithm.The experimental results show that the particle swarm optimization algorithm is superior to the artificial potential field method in convergence,accuracy and shortest path.Secondly,based on the knowledge of basic rough set,using various factors that affect the execution of soccer robot action in the dynamic environment establish the decision table for decision making system of soccer robot under the assumption of tactics.The redundancy factor is deleted by attribute reduction,so as to establish a most simple decision-making library for soccer robot to call.Although it is possible to reduce the decision-making system under the condition of complete rough set theory,there may be some data or some conditional attribute values are missing in the dynamic environment of the game.The decision making system will be formed in incomplete condition.Using decision table that application of basic rough set theory to reduce partial data integrity missing,the simplest decision table of the reduction is compared with the original decision table,the reliability of the extracted rules is analyzed.Under an improved rough set algorithm based on rough entropy and attribute importance degree is introduced to reduce the decision table,rule extraction and reliability analysis.