A Rolling Method for Complete Coverage Path Planning in Uncertain Environments

Abstract

Motion planning with obstacles avoidance in uncertain environments is an essential issue in robotics. Complete coverage path planning of a mobile robot requires the robot to pass through every area in the workspace with collision-free, which has many applications, e.g., various cleaning robots, painter robots, automated harvesters, land mine detectors and so on. A novel planning method integrating rolling windows and biologically inspired neural networks is proposed in this paper. The real-time environmental information can be represented by the dynamic activity landscape of the biological neural network. The rolling window approach is used to detect the local environments. Thus, a heuristic planning algorithm is performed on-line in rolling strategy. Three cases on complete coverage path planning in uncertain environments and the comparison of the proposed method and the planning approach based on the biologically inspired neural networks are studied by simulations. Simulation results show that the proposed method is capable of planning collision-free complete coverage robot motion path