An Improved Ant Colony Algorithm for UAV Path Planning in Uncertain Environment

Track planning for drones has been a common problem. In order to ensure that the UAV can fly long distance in accordance with the predetermined path, it is necessary to set calibration points on the flight path to correct the sensor errors of the UAV. This problem can be abstracted into a path planning problem and solved by ant colony algorithm. Considering the uncertainty of correction failure in these calibration points. This paper presents an improved ant colony algorithm and set up the "Enhanced pheromone volatilization strategy" to ensure that the UAV could reach the destination with the greatest possibility in this uncertain situation. We verify our algorithm on public data sets**. On data set 1, our algorithm has a 100% probability of reaching the destination, while the traditional ant colony algorithm has only a 61% probability of reaching the destination. On data set 2, our algorithm has a 56% probability of reaching the destination, while the traditional ant colony algorithm cannot find a path can reach the destination. The algorithm code*** in this paper is simple to implement, strong robustness, and can be extended to other scenarios.