Research On 3D Surveillance Trajectory Planning For Small UAV Based On Bio-inspired Algorithms

As one of the most important ways to deal with illegal drones,monitoring drones(MDr)are used to surville specific area to detect and combat the illegal drones.Due to the monitoring drones are usually powered by battery and are susceptible to enviro nment and its physical limitations,it is critical to plan a feasible and high-efficient surveillance trajecetory for monitoring drones.Based on the advantages of bio-inspired algorithm in solving complicated optimization problem,this thesis mainly focuses on MDr 3D surveillance trajectory planning in multi-MDrs and single MDr scenario,respectively.The main research contents and innovative works are summarized as follows:1.To solve the problems that the existing multi-UAVs path planning schemes lack of environmental adaptivity and cannot execute the surveillance task according to the surveillance area priority,we propose a multi-UAVs 3D surveillance Trajectory planning scheme.In this scheme,we first introduce the surveillance area importance(SAI)value to symbolize the surveillance area priority,and then combine it with other seven optimization indexes to design the multi-objective utility function which is utilized to determine the fitness of generated trajectories.After that,an event detection-based SAI value updating mechanism is proposed to enhance the environmental adaptivity.Finally,the particle swarm optimization(PSO)is used to derive the optimal trajectories.Simulation results validate that the trajectories generated by the proposed sche me can give priority to surveille the important areas first and the SAI value updating mechanism can help to trace the new events occurred in the operation area.2.Aiming at the problems of slow convergence speed and easiness to fall into loca l optimum of the classical ant colony algorithm(ACO)in solving 3D trajectory planning,this paper proposes an improved hybrid bio-inspired algorithm GAACO based on genetic algorithm(GA)and ant colony optimization algorithm(ACO).The proposed algorithm can dynamically adjust the number and quality of ants participating in pheromone updating by introducing a pheromone-updating threshold,therefore avoiding massive invalid search causing by the pheromone laid by the poor-quality ants.The GA operators and a local search technology are introduced to enhance the random Search capability of ant colony.To verify the performance of the proposed algorithm,we apply GAACO algorithm to the single monitoring drone's 3D surveillance trajectory planning.The simulation results validate that GAACO has stronger search ability than GA,ACO and PSO while maintaining faster convergence speed and stronger robustness.