机器人导航的自主选择搜索策略蜂群规划算法

为了减少机器人导航路径长度和路径规划时间,提出了基于自主选择搜索策略蜂群算法的规划方法。分析了人工蜂群算法原理,依据蜜蜂从自身认知、种群认知和其他个体认知等多种环境认知方式,对应给出了多种蜜源搜索方式;通过建立不同蜜源搜索方式的即时价值和后效价值模型,计算了蜜蜂选择不同蜜源搜索方式的概率,从而给出了蜜蜂对蜜源搜索方式的自主选择策略,在以上基础上提出了自主选择搜索策略蜂群算法。使用坐标旋转法将二维路径规划问题转化为一维,设计了两种环境下的导航路径规划仿真实验,在两种环境下自主选择搜索策略蜂群算法规划的路径长度均远远小于人工蜂群算法,且搜索到最优值的迭代次数也远远小于人工蜂群算法,充分证明了自主选择搜索策略蜂群算法在导航路径规划中的有效性。

Robot Navigation Path Planning Under Choose Searching Strategy Independently Bee Colony Algorithm

In order to lessen navigation path length of robot and path planning time-cost,path planning method based on choose searching strategy independently bee colony algorithm is proposed.Principle od artificial bee colony algorithm is analyzed.Depending on several environment cognation pattern,for example self-cognation,population cognation and other individual cognation,several honey source searching modes are given accordingly.Instant value model and feature value model of different honey source are built,choosing property of bee to different honey source searching mode is calculated.So that independently choose strategy of bee to honey source searching mode is given,based on the above,choose searching strategy independently bee colony algorithm is put forwarded.Two-dimension path planning method is transferred to one-dimension through rotation of coordinate.Navigation path planning simulation trial under two environment is designed,and path planned by choose searching strategy independently bee colony algorithm is obviously shorter than artificial bee colony algorithm under the two environment,and iteration time is also far less than artificial bee colony algorithm,which fully proves validity of choose searching strategy independently bee colony algorithm on navigation path planning.