Cooperative Obstacle Avoidance of Multiple Robotic Fishes Based on Grids Method

Abstract

To resolve the problem of obstacle avoidance and path planning of multiple robotic fishes, this paper proposes a new approach which uses collaboration mechanism based on grids method. The proposed approach splits the workspace of those robotic fishes using grids method, identifies each grid with serial number, designs the cooperative mechanism to avoid collision among these fishes, and plans the path for each fish. This method was applied in the obstacle avoidance competition of multiple robotic fishes which happening in a field with obstacle in it. The experimental results show the new method is more effective, can get more optimal path, and avoid the local minima issue which arises frequently in the A-star algorithm and genetic algorithm. It significantly improves the ability of the multiple robotic fishes system on the aspect of path planning and coordination.