基于混合结构的煤矿探测机器人导航控制策略

Navigation control strategy for rescue robots in colliery based on hybrid architecture

机器人的体系结构是系统研究的一个重要内容,由于传统系统模式对真实环境适应能力差、处理应激避障行为和直接规划行为过程复杂等一系列问题,笔者将反应式、反射式行为与慎思规划结合,提出一种混合式结构探测机器人导航控制策略。在处理应激避障行为和直接规划行为时,分别提出新的基于传感器的避障和规划,通过规划解决了复杂环境下的局部势能陷阱问题,在自行研制的救灾机器人CUMT-II上进行试验。结果证明,该控制结构行之有效。

System architecture is an important part in study of robot system. In view of poor adaptability of traditional mode to real surroundings, complex process of dealing with emergent obstacle avoidance and direct planning behavior, the paper integrated reactive behavior, reflecting behavior and deliberative planning, and proposed a kind of navigation control strategy for rescue robots based on hybrid architecture. While dealing with emergent obstacle avoidance and direct planning behavior, the paper proposed new obstacle avoidance and planning method based on sensors. The planning avoided local potential energy trap in complex surroundings. The proposed strategy proved feasible by the test on self-developed rescue robot CUMT-II.