For the characteristics of being unknown and time-varying of dynamic environment

a new method for mobile robot path planning is proposed in this paper. Firstly

in this method the environment model established by the grid method is convexized in order for the robot to avoid falling into U-shaped traps when moving along the planned path and thus to speed up the path planning. Secondly

a double-layer ant colony optimization (DACO) algorithm is proposed.In each iteration of DACO algorithm

a path is found firstly by the outer ACO algorithm

then on basis of which a small environment is constructed

and then the robot re-plans path by the inner ACO algorithm in the small environment; if the newly obtained path is better

then the global optimal path is updated and a new pheromone secondary update strategy proposed in this paper is executed. At last

three kinds of obstacle avoidance strategies are put forward according to the volume and speed of different dynamic obstacles in the environment. In the dynamic environment

the robot plans a global optimal path from the starting point to the destination with respect to the static environment via the DACO algorithm

then from the current starting point

the robot acquires the dynamic environment's information by its embedded sensor and implements waiting-for-obstacle avoidance strategy

collision avoidance strategy or rear-end collision avoidance strategy when necessary

and then the robot moves to next position as a new starting point. Simulation results show that the proposed can plan a safe and shortest path for mobile robot in real time under dynamic environment and is a practical and effective method for mobile robot path planning.