国家重点基础研究发展计划资助项目（2007CB311005）
为了解决清洁机器人完全覆盖路径规划中最大覆盖率和最小重复率的要求，在清洁机器人犁田式全局路径规划算法的基础上，提出了BP神经网络方法作为清洁机器人的局部路径规划。运用基于深度优先遍历的改进型BP神经网络算法，解决清洁机器人的清扫死区问题。仿真的结果表明所提出的BP神经网络方法和改进型BP神经网络算法能够解决清洁机器人在家庭内的完全覆盖路径规划问题。
To address the problem of planning complete coverage paths for cleaning robots that have the greatest coverage rates and the lowest recover rates, we propose a back propagation (BP) neural network algorithm for local path planning based on plowing global path planning. We also propose an improved BP neural network algorithm based on the deepfirst coverage algorithm for the issue of cleaning dead areas. Simulation results show that the proposed algorithm can plan complete paths for the robots.
王楷,石为人. BP神经网络在清洁机器人路径规划中的应用[J].重庆大学学报,2009,32(3):349-352.
复制
