摘要
针对智能体避障选取VFH及一系列改进方法时,原有直方图构造模型的复杂性问题以及大部分研究方法忽略了障碍物运动特性以及智能体自身大小对智能体避障的影响.提出一种新的直方图计算方法.该方法改进了原始的栅格计算模型,采用几何分析计算的方法,减少了智能体避障过程中的计算量;通过对动态物体进行研究,算法考虑了障碍物运动的特性以及智能体自身尺寸对避障所产生的影响.仿真实验表明,通过该算法构造的直方图能够作为智能体避障决策的基础,使得智能体能够顺利绕过动态障碍物,寻求优化的路径,快速追踪到静态目标和动态目标.
VFH and a series of improved method are selected for agent to avoid obstacles. Due to the complexity of the grid world model, it costs much time to compute the histogram. Also, many methods do not consider the feature that obstacles are dynamic and agent has its own size. This paper proposes a new method to compute histogram based on geometric analysis, which takes the move- ment of obstacles and the size of the agent into consideration. And, it reduces the amount of computation. Simulation results show that the histogram constructed by this method can be a foundation for agent obstacle avoidance decision; agent can search an optimal path to reach the static or dynamic target quickly.
出处
《小型微型计算机系统》
CSCD
北大核心
2014年第3期602-605,共4页
Journal of Chinese Computer Systems
基金
江苏省普通高校研究生科研创新计划项目(CXLX11_0203)资助
航空科学(2010ZC13012)资助
国防基础研究基金项目(Q072006C002-1)资助
