基于“蝇眼”网络模糊控制的装甲车避障系统研究

摘要

Armored car body weight,body length there such as congenital disadvantage,in the process of vehicle walking vision less than 180degrees,the body and the distance between the obstacles only by the human eye estimating,appear visual blind area,in turn and reverse often appear collision.This paper puts forward a based on the“fly's eye”network fuzzy control technology of armored vehicle obstacle avoid-ance system research,the armored car around the obstacles 360degree detection,real-time display bearing and distance,vehicle body of da-ta transmission using high stability of CAN bus technology,to realize the sensor information fusion based on neural network with fuzzy con-trol obstacle avoidance technology on armored vehicles to walk for intelligent obstacle avoidance control.In different environment of system test the obstacle avoidance method can make the obstacles bearing and distance measurement accuracy assurance in 95% and 98% above, have very strong practicability.%装甲车车身重，存在车体长宽等先天劣势，在车辆行走的过程中视野不到180度，车身与障碍物之间的距离只能通过人眼进行估测，驾驶员出现视觉盲区，在转弯与倒车时经常出现碰撞；提出一种基于“蝇眼”网络模糊控制技术的装甲车避障系统研究，对装甲车周围的障碍物进行360°的探测，实时显示方位与距离，车辆本体中的数据传输使用稳定性较高的 CAN 总线技术实现，对传感器的信息融合后使用基于神经网络模糊控制的避障技术对装甲车行走进行智能避障控制；在不同环境下的系统测试中本文的避障方法能够将障碍物的方位与距离测量准确率保证在95％与98％以上，具有很强的实用性。