机器人自主探索规划方法和系统设计

Robot Autonomous Exploration Planning Method and System Design

针对机器人在未知环境中存在效率低、即时性差、重复探索等问题,提出一种改进的自主探索方法。首先,对机器人局部视野范围进行优化,解决障碍物信息缺失时地图更新不完整的问题;在此基础上,将前沿边界点的信息增益作为奖励项,移动代价作为惩罚项构建非线性校用函数,对边界点进行合理的选择;最后,搭建了仿真及物理实验平台,实验证明所提方法优化了探索轨迹,有效提高了探索建图的效率。

An improved autonomous exploration method is proposed to solve the problems of low efficiency,poor immediacy and repeated exploration of robots in unknown environments.Firstly,the local field of vision of the robot is optimized to solve the problem of incomplete map update when the obstacle information is missing.On this basis,taking the information gain of the frontier boundary point as the reward term and the moving cost as the penalty term,a nonlinear correction function is constructed to reason-ably select the boundary point.Finally,the simulation platform is built and the efficiency is improved.