Complete Coverage Path Planning with Designated Starting and Target Locations Using a Neural Network Paradigm

摘要

Complete coverage path planning requires the robot path to cover all the unoccupied areas in the workspace. In some real world applications, the start and target locations have to be specified. It is desirable that a mobile robot is not only capable of starting from a specified location, planning a complete coverage path, and finally reaching a designated location after the cleaning is done, i.e., the robot performs a task by integrating the complete coverage path planning and the conventional pointto-point path planning. In this paper, a novel biologically inspired neural network paradigm is proposed for complete coverage path planning with designated starting and target locations in a nonstationary environment. The dynamics of each neuron in the discrete topologically organized neural network is characterized by a biologically inspired shunting equation. There are only local lateral connections among neurons. Therefore, the. computational complexity linearly depends upon the neural network size. The real-time robot path is autonomously generated from the dynamic activity landscape of the neural network and the previous robot location. No prior knowledge of the dynamic environment is needed. The main novelty of the proposed approach is that not only the robot is capable of autonomously performing the complete coverage path planning task with obstacle avoidance, but also it is able to start from a designated location and end in a designated target location.