基于克隆思维进化算法的机器人足球协作策略

针对足球机器人比赛系统的实时性要求,采用了一种克隆思维进化算法对足球机器人比赛系统的高层策略系统进行优化。克隆思维进化算法集免疫机制与进化机制于一体,在发挥思维进化算法优势的基础上增加了克隆(复制)、克隆重组、克隆变异和克隆选择等算子,既保持了种群的多样性,又提高了算法的收敛速度。足球比赛场上的瞬时信息作为抗原,待选策略作为抗体,二者均采用二进制编码方式。用克隆思维进化算法对抗体群进行优化,实验结果表明,采用该算法能快速找到最佳策略,简化了足球机器人决策系统,提高了决策效率。

Cooperation Strategy of Robot Soccer Based on Clonal Mind Evolutionary Algorithm

The clonal mind evolutionary algorithm was applied to the decision-making subsystem of robot soccer.The clonal mind evolutionary algorithm was a combination of the immune mechanism and evolution mechanism.Based on the maintenance of the advantage of the mind evolutionary algorithm,the clonal operator was added to the algorithm.It was beneficial to maintain the diversity of population and accelerate the convergence of the algorithm.The spot information of playground was regarded as antigen and the strategy was regarded as antibody.They were both described with binary code,the population of antibody was optimized according to the clonal mind evolutionary algorithm.It is helpful for robot to make decision effectively according to the environment.