Path Planning for Mobile Robots Based on an Improved Ant Colony Algorithm with Gaussian Distribution

Abstract Optimal path planning for mobile robots in complex environments has the problem of tangential collision with obstacles. This paper proposed an ant colony algorithm based on Gaussian distributed pheromone volatile mechanism (GD-ACO). The algorithm can reduce the collision probability with obstacles during the search navigation, and it increased the search efficiency and avoidance capability of mobile robot in complex environment path planning. Firstly, we established a grid environment model and introduced a Gaussian distribution pheromone volatility mechanism to perform optimal path planning. Secondly, we introduced right triangle constraints to eliminate the unqualified paths and reduce the complexity of the algorithm through a pheromone reward mechanism. Finally, by improving the heuristic information function distribution of the ant colony algorithm, it solved the algorithm cannot effectively converge to the optimal solution when encountering special obstacle distribution and achieved more efficient obstacle avoidance. Compared with other path planning algorithms, it can effectively generate a collision-free path. The simulation experiments show the improved ant colony algorithm has better obstacle avoidance capability and faster performance, and the proposed method can effectively find the optimal collision-free path through experimental verification.