基于产生式规则多传感器数据融合方法的移动机器人避障

Obstacle avoidance of mobile robot based on data fusion technology of production rule multi-sensor

在分析机器人多传感器数据融合技术的基础上,采用基于产生式规则的分布式多传感器数据融合方法,设计了面向多路不同类反射式传感器的数据融合规则库和移动机器人的避障决策规则库,并利用VisualC++编程实现了二级规则库及推理机制,在移动机器人平台上进行了实验。设计了不同场景的室内环境,结果机器人都能有效避免碰撞。实验表明多传感器信息融合能够在不提高单个传感器性能要求的情况下较好地改善移动机器人的自主避障能力。

On the base of the analysis of multi-sensor data fusion technology, using a multi-sensor data fusion method based on production rule, a data fusion rule base taking into consideration of multiple various types of reflective sensors and the obstacle avoidance rule base are built. The rules and the inference mechanism are implemented with Visual C＋＋. The experiments are carried out on a mobile robot with different layout in an indoor environment. The results show that the method can improve the obstacle avoidance capability of a mobile robot significantly without increasing the performance requirements for the single reflective sensors.