Obstacle Avoidance Of UAV Formation And Target Tracking Algorithms Based On Guidance

Along with the rapid development of the society,the application scope of unmanned aerial vehicles(UAV)has gradually shifted from enemy reconnaissance and surveillance,anti-missile,and counter-terrorism in the past,to urban management,agricultural plant protection,professional aerial photography,energy inspection,and other industries closely related to our life at present.The practicability of UAV is absolutely clear,especially in the areas where regular surveillance is required.Since the entry into the new century,universities and some scientific research institutions have also focused on theintelligent research on UAV system,including flight route planning,multi-UAV formation flight,target tracking,and etc.The research on UAV route planning,formation of obstacle avoidance for multi-UAV,and target tracking based on DJI Guidance module was carried out in this paper,with the specific content shown as follows:(1)Aiming at the problem of UAV track falling into local trap dead zone under the complex situation of continuous multiple obstacles.The fluid perturbation algorithm was improved,and the new flight route was developed by changing the streaming mode under the premise of not deviating from the main route to the greatest extent,so as to avoid the trap dead zone effectively.The deviation ratio of the improved fluid perturbation algorithm was optimized through the ordered iteration of the genetic algorithm,which could improve the selection efficiency of the offset coefficient and the success rate of the route planning.The simulation results showed that the proposed route is smooth and has a good obstacle avoidance performance.(2)The formation maintenance and obstacle avoidance problems of multi-UAV formation were studied.The topology diagram was established by using the information interaction of UAV,with the leader-following method as the basic structure of theformation pattern.The formation obstacle avoidance strategy based on consistency and improved fluid disturbance algorithm fusion and low number follower individual obstacle avoidance strategy was proposed to solve the problem of formation obstacle avoidance,formation transformation involving multiple UAVs and obstacle avoidance between UAVs.The simulation results demonstrated the effectiveness of this algorithm.(3)The improved CT tracking algorithm was transplanted to the visual system composed of Guidance and Manifold through the DJI SDK.In order to solve the defects of the fixed target tracking scale,the self-adaptive target scale method based on perceptual hashing was proposed,so that the CT algorithm could still effectively track the target object with the varying target sizes.Finally,the effectiveness of the proposed method was verified through experiment.