运用神经行为学原理提出一种模拟生物条件反射活动的机器人导航算法.对于环境感知,提出生物触角模型,因仅对特定区域信息进行处理,大大减小了计算量;对于运动控制,提出一种实用的改进Bug算法,机器人依靠生物触角感知环境刺激,实时激励触发相应的行走行为.最后在多种障碍物分布情形下进行算法对比验证.结果表明,机器人可以利用该算法实现未知环境下定目标点自主导航,具有转向次数少、绕行障碍物距离小及运动路径平滑等优点.该算法在自主移动机器人和自主牵引车辆中具有潜在应用价值.
This paper proposes a navigation algorithm of biomimetic conditioned reflex for mobile robots based on neuroethology principles. For environment txnception, a biological antenna model is proposed to delineate the interested sensing areas, and thus decrease the computational burden. For motion planning and control, according to a practical Bug algorithm proposed in this paper, the obtained environment stimulation is responded to generate the corresponding walking behavior. Experiments have been carried out in different conditions of obstacle distribution for algorithm verification. The results present its practicability, fewer turning number, smaller route length and smooth motion path. It can be expected that this algorithm will be of potential application in automated mobile robot(AMR) and automated guided vehicle(AGV).