Abstract: A new approach to complete coverage path planning for mobile robots,which integrates biologically inspired neural network,rolling window and heuristic searching,is presented.The local environment of mobile robot is modeled with Grossberg’s biological neural networks.The concept of rolling window is introduced into local path planning,and the local path planning goal in the rolling window can be determined by heuristic searching methods.The complete coverage path planning with dynamic obstacle avoidance for mobile robots can be efficiently implemented by the proposed method in uncertain dynamic environments.The effectiveness and feasiblity of the proposed method are illustrated by simulations.