机器人足球路径规划的一种改进算法

Improved Algorithm for Path Planning of Robot Soccer

在足球机器人系统路径规划问题的研究中,足球机器人系统工作的环境很复杂,既要配合本方机器人协同作战,还要对抗敌方机器人。针对传统的机器人路径规划算法过于复杂,同时没有充分考虑到足球机器人在比赛中实时性和对抗性,导致实时性差以及射门准确率低等问题,提出了一种在基本微粒群算法中加入交叉算子和变异算子的改进微粒群算法。种群中的所有微粒所经历过的最佳位置为当前全局最优位置。每个微粒都记忆当前自己所经历过的最优位置,将自己的最佳位置和全局最佳位置进行比较分析,修正自己的位置和速度。仿真结果显示改进微粒群算法比基本微粒群算法具有更优秀的搜索能力,实现了更快的收敛速度和更高的成功率,且对环境的变化有一定的适应性。

Working environment of the robot soccer system is very complex. It is necessary to cooperate with teammates and fight against the opposite robots. Traditional robot path planning algorithms were complex and not fully took into account real-time in the game of soccer robot and confrontational issues, leading to poor real-time and low shooting accurate. Crossover operator and mutation operator were added in the basic particle swarm algorithm in the paper. All the particles in the population have experienced the best position as the current global optimal position. Each particle remember experienced the optimal position of their own, compared the best position with the global best position, and modified their position and speed. Simulation results show that this algorithm is correct and effective with better search ability, fast convergence speed, higher success rate and good environmental adaptation.