Research On Obstacle Avoidance Algorithm Of UAV Power Line Patrol Based On Intelligent Video

The autonomous obstacle avoidance flight of UAV is of great significance for UAV.With the continuous expansion of the application scope of UAV,the application field of UAV is also increasing.In recent years,UAV also plays an increasingly important role in the field of power system,especially in the application of power line patrol.But at present,the application of UAV power system generally adopts artificial control,and the degree of intelligence is low.And the most important problem that UAV needs to overcome in power inspection is obstacle avoidance,and the core problem to be solved in obstacle avoidance is obstacle recognition and distance measurement between UAV and obstacle.Based on the theory of fuzzy quotient space,the method of designing dynamic granularity matrix is used to obtain the hierarchical structure of the image.Then,according to the hierarchical structure,the hybrid algorithm of group leader based on quantum heuristic is designed,and then the best granularity layer is calculated by combining the advantages of quantum evolution,group leader algorithm and logical mapping.After the optimal granularity layer randomly initializes the population,the corresponding The best two-dimensional segmentation threshold is obtained as the adaptive threshold of Canny edge detection,and the object detection and segmentation are realized according to the adaptive threshold of edge detection.Through the design of dynamic granularity matrix and hybrid algorithm,the fringe image is segmented and denoised.The designed algorithm has a good optimization effect on the detection and segmentation of obstacles in some poor environment.At the same time,through the parameter learning formula,the more accurate distance,angle and width information of obstacles are obtained.In this paper,the UAV is taken as the research object.Aiming at the route planning of UAV,the two-dimensional route planning method based on genetic algorithm and the three-dimensional route planning method based on artificial potential field are used.A chromosome coding method based on polar uniform change is proposed,which effectively decomposes the path between the starting point and the ending point,and solves the constraint problem of the path deflection angle.At the same time,the trajectory of the UAV in the two-dimensional plane is planned based on the two indexes of the best observation distance and the minimum total target distance,and the simulation results show that the UAV faces in the two-dimensional plane Both static and dynamic obstacles are avoided effectively.At the same time,based on the artificial potential field method in the three-dimensional plane,the mathematical model of gravitational potential field,repulsive potential field and obstacle point is established,and the three-dimensional trajectory planning path of UAV is obtained in the simulation experiment.