论文部分内容阅读
针对传统方法建立环境地图需要大量环境信息从而造成的存储搜索困难的问题,基于自组织特征映射图SOM提出了一种动态可增减自组织特征映射图DGPSOM算法,旨在用少量神经元来表征大量的环境信息,克服环境信息的存储问题,并在环境信息采集的过程中引入了信息熵的概念,通过计算不同方向上的信息熵来引导机器人探索未知环境,提高探索效率。然后根据DGPSOM算法建立的环境地图设计了一种智能A*搜索算法,搜索任意起点到终点的最优路径。仿真结果表明,建立的信息熵模型可以有效避免机器人重复访问同一区域,DGPSOM建立的环境地图可以实现以少量神经元表征大量环境信息,智能A*搜索算法可以完成机器人最优路径的搜索。
In view of the problem that the traditional method requires a large amount of environmental information to create an environmental map, the storage search is difficult. Based on the self-organizing feature map SOM, a dynamic self-organizing DGPSOM algorithm is proposed to characterize a small number of neurons A large amount of environmental information is obtained to overcome the storage problem of environmental information and the concept of information entropy is introduced in the process of environmental information collection. The information entropy in different directions is calculated to guide the robot to explore the unknown environment and improve the exploration efficiency. Then, an intelligent A * search algorithm is designed based on the environment map established by DGPSOM algorithm to search for the optimal path from any starting point to the ending point. The simulation results show that the established information entropy model can effectively avoid repeated robot visits to the same region. The environment map constructed by DGPSOM can represent a large amount of environmental information with a small number of neurons. The intelligent A * search algorithm can search the optimal path of the robot.