Research On UAV Path Planning Algorithms In Uncertain Environment

The path planning for UAV has become the key technique for UAV self-flying and auto-attacking. However, in the real flying environment, the threat resources that a UAV may encounter is becoming more and more complex and diversified. Hence, the study of path planning has become one of hot spots in UAV technology. In order to meet the demand for effective route of UAVs, the paper proposes modified ant colony optimization algorithm and modified artificial bee colony algorithm for UAV path planning.In view of the lack of slow convergence speed and easily falling into local optimum of the traditional ant colony optimization, this paper makes some improvements. Firstly, when state transition occurs in the traditional ant colony algorithm, the random probability selection strategy is used, which may lead to slow evolution because of the lack of guiding factor. Therefore, considering the selection strategy, the combination of random selection and pseudo random selection is introduced to improve the blindness of traditional ant colony optimization when searching, so as to solve the problem of slow evolution. Secondly, the pheromone evaporation factor ? remains unchanged in the basic ant colony optimization algorithm. When the flying environment becomes complex, the pheromone will decrease even close to zero on unsearched nodes. When ? is too large or too small, the convergence speed and the global search ability of the algorithm will be affected. Therefore, this paper sets upper and lower limits for pheromone evaporation factor ?, so as to improve the global search ability of the algorithm.In view of the slow convergence speed and earlier stagnation, this paper makes some improvements. Firstly, the chaotic sequence is introduced to initialize the nectar sources, it will ensure the randomness of initialization of the artificial bee colony algorithm and greatly enriched the diversity of nectar, producing large number of effective nectar sources, on the basis of which the better nectar sources can be chosen, thus the convergence speed can be speed up. Secondly, the selection mechanism is improved to avoid the phenomenon of earlier stagnation.Finally, with the platform of Matlab and VC++6.0, the experiments are carried out with modified ant colony algorithm and modified artificial bee algorithm respectively, proving that the modified algorithms are correct and valid. The experiments show that modified algorithms acquired better performance in both time consuming and path distance.